Optimal pathway control method for regional vegetation carbon storage based on multi-scenario simulation

Abstract

The present invention relates to the field of regional carbon storage prediction methods, and specifically to a method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation, determining the scope of a research area, collecting multi-source data sets for regional vegetation carbon storage simulation and multi-scenario simulation; preprocessing the multi-source data and building a database; correcting the maximum light utilization rate parameter; comparing planned data and multi-period land use data, building a variety of scenarios consisting of future climate change and land use change, and forming 12 regional vegetation carbon storage estimation scenarios; calculating the regional vegetation carbon storage under each scenario based on the combined characteristic data of the regional vegetation carbon storage estimation scenarios; estimating the changing trend of the regional vegetation carbon storage for each scenario; setting the constraints and optimization targets of the optimization path regulation, and optimizing and regulating the multi-path vegetation carbon storage strategy. The present invention can more scientifically and accurately improve the national land space pattern optimization path, regulation and management strategy of regional vegetation carbon storage.

Description

Technical Field

[0001] The present invention relates to the field of regional carbon storage prediction methods, and in particular to a regional vegetation carbon storage optimization path regulation method based on multi-scenario simulation. Background Art

[0002] In recent years, due to the influence of climate and human activities, the spatial pattern of regional vegetation carbon storage has undergone drastic changes. We are faced with the problem of how to maintain and enhance the regional vegetation carbon storage and carbon sink potential. The refinement and scientific management of the spatial distribution pattern of regional vegetation carbon storage and carbon sink potential are still worth discussing. Therefore, maintaining and enhancing regional vegetation carbon storage and carbon sink potential and enhancing the regional vegetation carbon fixation and sink function are important means to slow down the rise in atmospheric carbon dioxide concentration and global warming.

[0003] Under future climate change scenarios (SSP1-2.6 low forcing scenario, SSP2-4.5 medium forcing scenario, SSP5-8.5 high forcing scenario, etc.) and land use change scenarios (natural development scenario, economic development scenario, ecological protection scenario, sustainable development scenario, etc.), how to adopt scientific and reasonable approaches to implement protection and restoration of regional vegetation, precise greening, scientific management and rational utilization, and regulate the optimal layout strategy of regional vegetation at multiple temporal and spatial scales is a key technical issue that restricts the maintenance and improvement of regional vegetation carbon storage. Summary of the invention

[0004] The present invention aims to provide a method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation, so as to accurately and scientifically plan the maintenance and improvement of regional vegetation carbon storage.

[0005] The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation in this scheme includes the following steps:

[0006] Step 1, determine the scope of the study area, collect relevant multi-source data sets required for regional vegetation carbon storage simulation and multi-scenario simulation, and the multi-source data include regional vegetation carbon storage simulation data and multi-scenario simulation data;

[0007] Step 2: preprocess the vegetation data, meteorological data, land use data and climate scenario data in the multi-source data and build a database;

[0008] Also includes:

[0009] Step 3, modify the maximum light utilization parameters in the CASA model applicable to different ecosystems / land uses in the study area;

[0010] Step 4: Compare the planned data with multi-period land use data to construct multiple scenarios of future climate change and land use change, and form 12 regional vegetation carbon storage estimation scenarios;

[0011] Step 5, based on the combined characteristic data of 12 regional vegetation carbon storage estimation scenarios, the CASA model is used to calculate the regional vegetation carbon storage VCS under each scenario;

[0012] Step 6, based on the regional vegetation carbon storage VCS of 12 regional vegetation carbon storage estimation scenarios, the trend analysis method is used to estimate the change trend of regional vegetation carbon storage;

[0013] Step 7, set the maximum values ​​of vegetation carbon storage and flux as the constraints for optimizing the path control, match different development paths with optimization goals according to scenarios, and optimize and control the vegetation carbon storage of regional land use quantity structure, spatial layout, and management strategy in multiple paths.

[0014] The beneficial effects of this program are:

[0015] By considering the optimal targets of vegetation carbon storage and flux under the combined effects of 12 future changing environments, the spatial distribution pattern of regional vegetation to maintain and enhance its carbon storage and carbon sink potential is managed in a refined and scientific manner. A scenario of future climate change combined with land use change is constructed to accurately predict the maximum potential for vegetation carbon sequestration and carbon sink enhancement in the study area in the future. Constraints are set for regulation of the optimal path with the maximum vegetation carbon storage potential, and technical paths for improving the vegetation carbon sequestration capacity under different spatial land use patterns in the region are guided, thereby scientifically regulating and formulating the spatial distribution pattern and management policies for maintaining and enhancing the regional vegetation carbon storage and carbon sink potential.

[0016] Further, in step 1, the regional vegetation carbon storage simulation data includes vegetation, meteorological, and light energy utilization efficiency parameter data required for estimating vegetation net primary productivity using the CASA model, the vegetation data includes the normalized vegetation index on a monthly scale, the meteorological data includes precipitation, temperature, and solar radiation, and the light energy utilization efficiency parameters include land use / cover type and maximum light energy utilization efficiency parameter data;

[0017] The multi-scenario simulation data include climate scenario data of the coupled model comparison plan and multi-period land use data, and the climate scenario data include monthly-scale normalized difference vegetation index, precipitation, temperature, and solar radiation scenario data under low forcing scenarios, medium forcing scenarios, and high forcing scenarios.

[0018] The beneficial effect is that through multi-source data sources and various types of data settings, the accuracy and comprehensiveness of the data sources used for analysis can be improved.

[0019] Further, in step 2, the preprocessing includes: by comparing the projection information and pixel size of multi-source raster data of vegetation data, meteorological data, and land use data, based on the minimum raster pixel, using a series of operations such as format conversion, clipping, projection conversion, and resampling for preprocessing;

[0020] By establishing the classification standard of land use according to the land resources and their utilization attributes, the land use types in the study area are divided into 6 categories, including cultivated land, forest land, grassland, water area, construction land, and unused land for pre-processing;

[0021] Through the downscaling method, the multi-scenario simulation data of the large-scale, low-resolution global climate model output information are converted into small-scale, high-resolution regional ground climate change information. The simulation values ​​preprocessed by the downscaling method are compared with the historical observation data. The root mean square error, correlation coefficient, relative deviation coefficient and Nash-Sutcliffe model efficiency coefficient between the simulation values ​​and the observation values ​​are used to examine the simulation accuracy of temperature and precipitation. Random points are selected from space for correlation test the reliability of the simulation results.

[0022] The beneficial effects are: the data format is unified and corrected to ensure the consistency of the input data's geographic spatial range, projection coordinates, spatial resolution, row and column numbers, and simulation time period.

[0023] Furthermore, in step 4, climate scenarios are constructed, and typical future climate change scenarios of low forcing scenarios, medium forcing scenarios, and high forcing scenarios are obtained based on coupled model comparison program data;

[0024] Based on the land use type data, the Markov model and PLUS model are used to predict the future land use change scenarios. According to the regional land use conversion law and combined with the future development plan of the region, four scenarios are set: natural development scenario, economic development scenario, ecological protection scenario, and sustainable development scenario.

[0025] Based on four scenarios, the shared socio-economic development pathway and the typical concentration pathway combination scenario were coupled to construct 12 future climate change scenarios considering ground conditions.

[0026] The beneficial effect is that the constructed future climate change scenarios can ensure complete coverage of known climate conditions.

[0027] Furthermore, in step 4, the 12 regional vegetation carbon stock estimation scenarios include a low forcing scenario superimposed on a natural development scenario, a low forcing scenario superimposed on an economic development scenario, a low forcing scenario superimposed on an ecological protection scenario, a low forcing scenario superimposed on a sustainable development scenario, a medium forcing scenario superimposed on a natural development scenario, a medium forcing scenario superimposed on an economic development scenario, a medium forcing scenario superimposed on an ecological protection scenario, a medium forcing scenario superimposed on a sustainable development scenario, a high forcing scenario superimposed on a natural development scenario, a high forcing scenario superimposed on an economic development scenario, a high forcing scenario superimposed on an ecological protection scenario, and a high forcing scenario superimposed on a sustainable development scenario.

[0028] The beneficial effect is that the superposition setting of regional vegetation carbon stock estimation scenarios can ensure that the designed estimation scenarios are consistent with the actual variability scenarios, making the estimation results more accurate.

[0029] Furthermore, in step 5, based on NDVI, temperature, precipitation, solar radiation, and land use data under 12 regional vegetation carbon storage estimation scenarios, the net primary productivity NPP of vegetation at a monthly scale within a preset time period in the region simulated by the CASA model;

[0030] The growing biomass density and carbon storage of regional vegetation were calculated using the CASA model and vegetation net primary productivity (NPP).

[0031] The beneficial effect is that the calculation of various parameters can accurately quantify the carbon storage situation.

[0032] Furthermore, in step 6, the Theil-Sen Median trend method and the Mann-Kendall test method are combined to identify the long-term series change trend characteristics of regional carbon storage within a set time period, and five change trend results are obtained: significant degradation, slight degradation, basically stable, slight recovery, and significant recovery.

[0033] The beneficial effect is that by setting the trend characteristics of carbon storage changes in a certain area in the future, the corresponding carbon storage changes can be intuitively and quickly represented.

[0034] Furthermore, in step 6, the Theil-Sen Median trend method is used to analyze the changing trend of regional vegetation carbon storage under 12 scenarios. The Theil-Sen Median trend calculation formula is:

[0035]

[0036] Among them, median is the median function, x j 、x i Represent the data of the jth and ith time series respectively;

[0037] When Slope is greater than 0, it indicates an increasing trend, otherwise it indicates a decreasing trend.

[0038] The beneficial effects are: Theil-Sen Median trend method can effectively reflect the degree of change and trend of factors, has no specific requirements for the data distribution type, does not require the data to follow a certain distribution, and can have a strong resistance to data errors.

[0039] Furthermore, in step 6, the Mann-Kendall test method performs a trend test to identify the long-term series change trend characteristics of regional carbon storage within a set time period, and the detected statistic S is expressed as:

[0040]

[0041] Where S is a normal distribution with a mean of 0 and a variance of Var(S)=n(n-1)(2n+5) / 18; time series data (X1,…,X n ) is a sample of n independent random variables with the same distribution. For all i, j≤n, and when i≠j, the distributions of Xi and Xj are different;

[0042] When n>10, the standard normal system variable calculation formula is:

[0043]

[0044] Given a confidence level α, if |Z|≥Z 1-α / 2 , the null hypothesis is not accepted, the confidence level is set to α=0.05, when |Z|≤1.96, the null hypothesis is accepted, when Z>1.96, the sequence has a significant upward trend, when Z<-7.96, the sequence has a significant downward trend.

[0045] The beneficial effect is that the test method can test the accuracy of the changing trend determined by the trend method.

[0046] Further, in step 7, the maximum value of vegetation carbon storage and flux is set as the multi-objective optimization model. In the multi-objective optimization model, 12 scenarios are set, and the functions are realized as 2. The multi-objective optimization model has a total of 24 decision variables. The multi-objective optimization model has 2 objective functions, namely, the maximum vegetation carbon storage target f1(x) and the maximum vegetation carbon storage flux target f2(x). Then the multi-objective optimization model is expressed as:

[0047] obj.F(x)=opt{f1(x),f2(x)}:

[0048] The vegetation carbon stock inventory target is expressed as:

[0049]

[0050] Where q and m represent the qth land use type and the number of land use types in the region, respectively; VCD q Represents the carbon density of the qth land use type in the region, in t / hm 2 ;x q Represents the total area of ​​the qth land use type in the region, in hm 2 ;

[0051] The vegetation carbon storage flux target is expressed as:

[0052]

[0053] Where: q and m represent the qth land use type and the number of land use types in the region respectively; e q represents the vegetation carbon storage change rate of the qth land use type in the region; x q Represents the total area of ​​the qth land use type in the region, in hm 2 ;

[0054] Spatial constraints and area constraints are used as constraints for optimizing path regulation. The spatial constraint is that ecological land within the region cannot be converted into other land types, and the area constraint is that the total area of ​​each type of land is constant to the area of ​​the study area.

[0055] The non-dominated sorting genetic algorithm II is used to solve the multi-objective optimization model;

[0056] Through multi-scenario comparative analysis, multi-objective optimization and multi-criteria decision-making models are combined with scenario analysis methods to optimize the national land space layout with the goal of reducing emissions and increasing sinks in terrestrial ecosystems.

[0057] The beneficial effect is that the optimization of the national land space layout can analyze the corresponding carbon storage situation in advance, so as to make corresponding arrangements according to the target designed carbon storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of an embodiment of the method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to the present invention. DETAILED DESCRIPTION

[0059] The following is further explained in detail through specific implementation methods.

[0060] Example

[0061] The optimization path control method of regional vegetation carbon storage based on multi-scenario simulation, such as Figure 1 As shown, the following steps are included:

[0062] Step 1: Determine the scope of the study area and collect relevant multi-source data sets required for regional vegetation carbon storage simulation and multi-scenario simulation. The multi-source data include regional vegetation carbon storage simulation data and multi-scenario simulation data. The specific analysis is based on the historical period of 2000-2020 and the future period of 2025-2050 for data and targets.

[0063] The regional vegetation carbon storage simulation data include the vegetation, meteorological, and light energy utilization efficiency parameter data required for estimating the net primary productivity of vegetation using the CASA model. The vegetation data include the normalized difference vegetation index (NDVI) on a monthly scale. The meteorological data include precipitation, temperature, and solar radiation. The NDVI data comes from the GIMMS AVHRR NDVI3g data (1982-2022). The meteorological data of temperature, precipitation, and solar radiation come from the National Meteorological Science Data Center. The light energy utilization efficiency parameters include land use / cover type and maximum light energy utilization efficiency parameter data.

[0064] The multi-scenario simulation data include the climate scenario data and multi-period land use data of the Sixth Coupled Model Intercomparison Project (CMIP6). The climate scenario data include the monthly scale normalized difference vegetation index, precipitation, temperature, and solar radiation scenario data under the SSP1-2.6 low forcing scenario, SSP2-4.5 medium forcing scenario, and SSP5-8.5 high forcing scenario. The time scale of these scenario data is 2000 to 2050.

[0065] Step 2: Preprocess the vegetation data, meteorological data, land use data and climate scenario data in the multi-source data and build a database. The preprocessing includes: by comparing the projection information and pixel size of the multi-source raster data of vegetation data, meteorological data and land use data, based on the minimum raster pixel, using ArcGIS software or R language and other tools, using a series of operations such as format conversion, clipping, projection conversion and resampling for preprocessing.

[0066] Data preprocessing is achieved through batch processing through R code programming. The original data in *.nc format are first converted into *.tif format, and a coordinate system is assigned. The resolution is unified to 30m based on the minimum grid pixel to ensure consistency with the resolution, projection coordinates, and row and column numbers of NDVI, temperature, precipitation, solar radiation, land use and other data. The monthly raster dataset of the study area is cropped according to the boundary of the study area.

[0067] By establishing a land use classification standard based on land resources and their utilization attributes, the land use types in the study area are divided into six categories, including cultivated land, forest land, grassland, water area, construction land, and unused land for preprocessing. The specific classification is shown in Table 1.

[0068] Table 1 Classification system standards for multi-period land use remote sensing monitoring datasets

[0069]

[0070]

[0071] The multi-scenario simulation data of the large-scale, low-resolution global climate model output information is converted into small-scale, high-resolution regional ground climate change information through the downscaling method, and the simulation values ​​preprocessed by the downscaling method are compared with the historical observation data. The root mean square error, correlation coefficient, relative deviation coefficient, and Nash-Sutcliffe model efficiency coefficient between the simulation values ​​and the observation values ​​are used to examine the simulation ability of temperature and precipitation, and random points are selected from space for correlation test to verify the reliability of the simulation results. The downscaling method uses bilinear interpolation to uniformly interpolate the climate model output data of different resolutions to the corresponding stations of the same longitude and latitude, and the interpolated data are bias-corrected and compared with the historical observation data. The root mean square error, correlation coefficient, relative deviation coefficient, and Nash-Sutcliffe model efficiency coefficient between the simulation values ​​and the observation values ​​are used to examine the simulation ability of temperature and precipitation, and random points are selected from space for correlation test to further illustrate the reliability of the simulation results.

[0072] Specifically, the measured data from meteorological stations and field surveys are used as the basis for evaluating the "true value". The four evaluation indicators, namely, correlation coefficient (CC), root mean square error (RMSE), relative bias coefficient (PBIAS), and Nash-Sutcliffe model efficiency coefficient (NSE), are used to evaluate the accuracy, reliability, and applicability of CMIP6 data in the study area. The specific calculation formula is as follows:

[0073]

[0074] Where: Si represents the evaluation value of the i-th CMIP6 data and simulation data; O i represents the true value of measured data from the ith meteorological station, field survey, etc.; n represents the number of evaluation data samples; Represent the average values ​​of sample evaluation data and actual observation data respectively.

[0075] CC is mainly used to analyze the degree of correlation and consistency between grid products and simulated data and measured data. The value range is [-1,1]. Positive and negative data indicate the direction of correlation. The closer the value is to 1, the stronger the correlation. RMSE describes the error between grid products and simulated data and measured data. The value range is non-negative real numbers. The closer the value is to 0, the higher the accuracy and the better the applicability of grid products and simulated data. PBIAS is used to describe the relative deviation between grid products and simulated data and measured data. The positive and negative represent the direction of error and deviation, respectively indicating that the evaluation data is overestimated or underestimated than the measured value. The closer the value is to 0, the higher the accuracy and the better the applicability of grid products and simulated data. NSE is usually used to verify the quality of hydrological model simulation results to analyze the reliability of grid products and simulated data compared with measured data. The NSE value range is from negative infinity to 1, and is divided into the following three cases: if NSE is close to 1, it means that the data quality of the grid products and simulated data is good and the credibility is high; if NSE is close to 0, it means that the evaluation value of the grid products and simulated data is close to the average level of the observed real data, and the overall result is relatively credible, but the error is relatively large; if NSE is much less than 0, it means that the grid products and simulated data are unreliable.

[0076] Since climate scenario data often have a low resolution, downscaling can make up for the limitations of global climate models in predicting regional climate change, and convert the large-scale climate information provided by global climate models into regional-scale climate information (such as temperature, precipitation, etc.), thereby achieving regional climate prediction. The spatial resolution and projection information of downscaling are consistent with multi-source raster data such as vegetation data (NDVI), meteorological data (temperature, precipitation, solar radiation, etc.), and land use data.

[0077] Step 3: Modify the maximum light utilization rate parameters in the CASA model that are applicable to different ecosystems / land uses in the study area. The NPP estimation in the CASA model is mainly calculated through the photosynthetically active radiation (APAR) absorbed by plants and the actual light energy utilization efficiency (ε). The formula for estimating vegetation net primary productivity is as follows:

[0078] NPP(x,t)=APAR(x,t)*ε(x,t);

[0079] Where APAR(x, t) represents the photosynthetically active radiation absorbed by pixel x at time t, in units of g C / m 2 ; ε(x, t) represents the actual light energy utilization rate of a single pixel x in month t, in g C / MJ;

[0080] APAR (x, t) is determined by the ratio of the solar effective radiation that can be absorbed by the vegetation to the incident photosynthetically active radiation absorbed by the vegetation. The formula is as follows:

[0081] APAR(x,t)=S0L(x,t)*FPAR(x,t)*0.5;

[0082] Where SOL(x, t) represents the total solar radiation at a single pixel x in month t, in units of g C / m 2 ; FPAR(x, t) represents the absorption ratio of the vegetation layer to the incident photosynthetically active radiation, and the constant 0.5 represents the ratio of the solar effective radiation (wavelength 0.4-0.7 μm) that can be used by vegetation to the total solar radiation.

[0083] Light energy utilization efficiency is the ratio of the chemical potential contained in the dry matter produced per unit area in a certain period to the photosynthetically active radiation energy projected onto the area at the same time. Environmental factors such as temperature, soil moisture conditions, and atmospheric water vapor pressure difference will regulate the NPP of vegetation by affecting the photosynthetic capacity of vegetation. The calculation formula is as follows:

[0084] ε(x,t)=T ε1 (x,t)*T ε2 (x,t)*W ε (x,t)*ε max ;

[0085] Where, T ε1 (x, t) and T ε2 (x, t) represents the stress effect of low temperature and high temperature on light energy utilization; W ε (x, t) is the water stress coefficient, reflecting the effect of water conditions on light energy utilization; ε max It is the maximum light energy utilization under ideal conditions, and its unit is gC / MJ.

[0086] In order to avoid the vegetation light energy utilization rate ε max The error brings uncertainty to the estimation of regional vegetation NPP. We use measured data to evaluate and compare two commonly used NPP accuracy evaluation methods based on other research results to re-determine the vegetation light energy utilization efficiency ε of different ecosystems in the study area. max . Maximum light energy utilization ε max The values ​​of refer to Zhu Wenquan, Pan Yaozhong, Zhang Jinshui. Remote sensing estimation of net primary productivity of terrestrial vegetation in China [J]. Acta Plant Ecologica Sinica, 2007, (03): 413-424. The values ​​used in the study of simulating the light energy utilization efficiency of typical vegetation in China. The other parameters are taken as the global monthly average maximum light utilization efficiency of 0.389 g C / MJ.

[0087] Step 4, compare the planned data with multi-period land use data, construct multiple scenarios of future climate change and land use change composed of SSP1-2.6 low forcing scenario, SSP2-4.5 medium forcing scenario, SSP5-8.5 high forcing scenario, and form 12 regional vegetation carbon stock estimation scenarios, including low forcing scenario superimposed on natural development scenario, low forcing scenario superimposed on economic development scenario, low forcing scenario superimposed on ecological protection scenario, low forcing scenario superimposed on sustainable development scenario, medium forcing scenario superimposed on natural development scenario, medium forcing scenario superimposed on economic development scenario, medium forcing scenario superimposed on ecological protection scenario, medium forcing scenario superimposed on sustainable development scenario, high forcing scenario superimposed on natural development scenario, high forcing scenario superimposed on economic development scenario, high forcing scenario superimposed on ecological protection scenario, and high forcing scenario superimposed on sustainable development scenario. The description of future climate change scenarios is shown in Table 2.

[0088] Table 2 Meaning of description of future climate change scenarios

[0089]

[0090] Based on the land use type data, the Markov model and PLUS model are used to predict the future land use change scenarios. Based on high-precision remote sensing images such as the Landsat series of satellites, a high-precision multi-period land use data set of the region is obtained through interpretation, and the Markov model and PLUS model are coupled to accurately simulate and predict the future land use spatial distribution scenarios. According to the regional land use conversion law and combined with the future development plan of the region, four scenarios are set: natural development scenario, economic development scenario, ecological protection scenario, and sustainable development scenario, as shown in Table 3.

[0091] Table 3 Future development trends of the region and the implications of the four land use change scenarios set by policy

[0092]

[0093] The study set up four scenarios for land use in 2050, including natural development scenario, economic development scenario, ecological protection scenario, and sustainable development scenario. Based on the existing research on the response scenario simulation of land use change in typical karst areas to the value of ecosystem services, the following conversion probability was designed in combination with the actual situation of the study area.

[0094] ① Natural development scenario: The future land use quantity is predicted according to the MarKov model of the PLUS model, and the conversion probability between land types is predicted.

[0095] ② Economic development scenario. The MarKov model of the PLUS model is used to predict the future land use quantity. Based on the natural development trend, the probability of converting cultivated land to other land types is reduced by 50%, and the probability of converting water area, grassland, and unused land to cultivated land is increased by 20%, 30%, and 50%, respectively. The probability of other land types maintaining natural development is maintained.

[0096] ③ Ecological protection scenario. The MarKov model of the PLUS model is used to predict the future land use quantity. Based on the natural development trend, the probability of water area, grassland, and construction land being transferred to forest land is increased by 20%, 20%, and 10%, respectively. The probability of forest land, cultivated land, water area, and grassland being converted to construction land is reduced by 50%, 30%, 20%, and 30%, respectively. The probability of forest land being transferred to grassland is reduced by 50%.

[0097] ④ Sustainable development scenario. The MarKov model of the PLUS model is used to predict the future land use quantity. Based on the natural development trend, the probability of converting cultivated land to other land types (except forest land) is reduced by 50%, the probability of converting grassland, forest land, and water area to construction land is reduced by 25%, 50%, and 25%, respectively, the probability of converting forest land to other land use is reduced by 50%, and the probability of converting water area, grassland, and construction land to forest land is increased by 30%, 30%, and 20%, respectively.

[0098] On this basis, the combined scenarios of shared socioeconomic development pathways (SSPs) and representative concentration pathways (RCPs) are coupled to construct a variety of future climate change scenarios considering ground conditions. The specific characteristics of the 12 scenario combinations are shown in Table 4.

[0099] Table 4 Characteristics of regional vegetation carbon storage estimation scenario combinations

[0100]

[0101] Step 5: Based on the combined characteristic data of 12 regional vegetation carbon stock estimation scenarios, the CASA model is used to calculate the regional vegetation carbon stock VCS under each scenario. Based on the NDVI, temperature, precipitation, solar radiation, and land use data under the 12 regional vegetation carbon stock estimation scenarios, the monthly vegetation net primary productivity NPP in the regional preset time period simulated by the CASA model is calculated. The preset time period is 2000-2050.

[0102] The CASA model and the vegetation net primary productivity (NPP) are used to calculate the biomass density and carbon storage of regional vegetation. Vegetation carbon storage (VCS) is the amount of carbon in the vegetation layer carbon pool (forest, land, etc.), usually measured in tons. Vegetation net primary productivity (NPP) is considered an important link in the carbon cycle and carbon balance of terrestrial ecosystems, and is also a major factor and important indicator for determining ecosystem carbon sinks and regulating ecological processes. Vegetation carbon storage includes the carbon storage of all vegetation types in the region, and the calculation method is as follows:

[0103]

[0104] Where: VCS represents the total carbon storage of regional vegetation, unit t; i and n represent the number of vegetation types and species of the i-th vegetation type in the region, respectively; VCD i represents the carbon density of the i-th vegetation type in the region (t / hm2); S i Represents the total area of ​​the i-th vegetation type in the region, in hm 2 .

[0105] Vegetation carbon density indicates the amount of carbon contained in vegetation per unit area. The calculation method for vegetation carbon density is:

[0106] CDV i =NPP i ×CF i ;

[0107] In the formula, CDV i Indicates the carbon density of the i-th vegetation type in the region, in t / hm 2 ; NPP i It represents the net primary productivity of vegetation per unit area of ​​the i-th vegetation type in the region, in t / hm 2 CF i The conversion factor representing the net primary productivity of vegetation type i in region i is usually taken as 0.565.

[0108] Step 6: Based on the regional vegetation carbon stock VCS of 12 regional vegetation carbon stock estimation scenarios, the trend analysis method is used to estimate the change trend of regional vegetation carbon stock. The Theil-Sen Median trend method and the Mann-Kendall test method are combined to identify the long-term series change trend characteristics of regional carbon stock within the set time period. The Slope and Z values ​​are combined for judgment. As shown in Table 5, five change trend results are obtained: significant degradation, slight degradation, basically stable, slight recovery, and significant recovery. Significant degradation can also be a significant decrease, slight degradation can also be a non-significant decrease, slight recovery can be a non-significant increase, and significant recovery can be a significant increase.

[0109] Table 5 Trend analysis classification standards

[0110] Slope Z value Trend Changes Slope ≥ 0.0001 Z>1.96 Significant increase (or significant recovery) Slope ≥ 0.0001 |Z|≤1.96 No significant increase (or slight recovery) -0.0001<Slope<0.0001 |Z|≤1.96 Basic stability Slope≤-0.0001 |Z|≤1.96 Non-significant decrease (or slight degradation) Slope≤-0.0001 Z<-1.96 Significantly reduced (or significantly degraded)

[0111] By judging the various numerical values ​​in the table, we can form a standard for accurately judging the significance of trend analysis and analyze the changing trend of regional vegetation carbon storage.

[0112] Among them: Theil-Sen Median trend method is used to analyze the changing trend of regional vegetation carbon storage under 12 scenarios. The Theil-Sen Median trend calculation formula is:

[0113]

[0114] Among them, median is the median function, x j 、x i Represent the data of the jth and ith time series respectively;

[0115] When Slope is greater than 0, it indicates an increasing trend, otherwise it indicates a decreasing trend.

[0116] The Mann-Kendall test method is used to perform trend tests to identify the long-term series change trend characteristics of regional carbon storage within a set time period. The test statistic S is expressed as:

[0117]

[0118] Where S is a normal distribution with a mean of 0 and a variance of Var(S)=n(n-1)(2n+5) / 18; time series data (X1, ..., X n ) is a sample of n independent random variables with the same distribution. For all i, j≤n, and when i≠j, the distributions of Xi and Xj are different;

[0119] When n>10, the standard normal system variable calculation formula is:

[0120]

[0121] Given a confidence level α, if |Z|≥Z 1-α / 2 , the null hypothesis is not accepted, the confidence level is set to α=0.05, when |Z|≤1.96, the null hypothesis is accepted, when Z>1.96, the sequence has a significant upward trend, when Z<-1.96, the sequence has a significant downward trend.

[0122] Step 7, set the maximum values ​​of vegetation carbon storage and flux as the constraints for optimizing the path control, match different development paths with optimization goals according to scenarios, and optimize and control the multiple paths of vegetation carbon storage in the quantitative structure, spatial layout, and management strategies of regional land use.

[0123] Multi-objective optimization refers to finding a solution that can achieve the best balance between multiple objectives when there are multiple decision variables and multiple objective functions. Optimization models include traditional optimization methods, bionic intelligent optimization models, and optimization decisions based on game theory. Traditional multi-objective optimization methods include the main objective method, hierarchical sequence method, weighted method, ideal point method, etc., which are characterized by converting multi-objective optimization problems into single-objective optimization problems for solution. Evolutionary algorithms such as genetic algorithms and artificial immunity, as well as ant colonies and particle swarms are commonly used bionic intelligent algorithms, which are widely used in complex optimization problems with high dimensions, nonlinearity, and multiple solution sets. The equilibrium strategy in game theory, the cooperation and confrontation of game participants, etc. can effectively consider the diversity of human decision-making and values ​​to model and solve corresponding complex problems, while reducing the loss of solution efficiency and solution quality.

[0124] The multi-objective optimization model can achieve the prediction of spatial quantity structure by setting multi-objective functions, based on the constraint data and change rules of decision variables, and can find a better combination of decision variables from the multi-objective system. Because the prediction results of this method are more scientific and reasonable, it can also integrate multiple objectives for coordinated optimization, making it highly applicable in national land space optimization.

[0125] The multi-objective optimization model based on vegetation carbon storage and flux can realize the prediction of spatial quantity structure by setting multi-objective functions, based on the constraint data and change law of decision variables, and can find a better combination of decision variables from the multi-objective system. Because the prediction results of this method are relatively scientific and reasonable, it can also integrate multiple objectives for collaborative optimization, making it highly applicable when optimizing national land space. The maximum value is used to construct a multi-objective optimization model. In the multi-objective optimization model, 12 scenarios are set, and the function is realized as 2. The multi-objective optimization model has a total of 24 decision variables. The multi-objective optimization model has 2 objective functions, namely the maximum vegetation carbon storage target f1(x) and the maximum vegetation carbon storage flux target f2(x). The multi-objective optimization model is expressed as:

[0126] obj.F(x)=opt{f1(x),f2(x)}:

[0127] The vegetation carbon stock inventory target is expressed as:

[0128]

[0129] In the formula, q and represent the qth land use type and the number of land use types in the region respectively; VCD q Represents the carbon density of the qth land use type in the region, in t / hm 2 ;x q Represents the total area of ​​the qth land use type in the region, in hm 2 ;

[0130] The vegetation carbon storage flux target is expressed as:

[0131]

[0132] Where: q and m represent the qth land use type and the number of land use types in the region respectively; slope q It represents the vegetation carbon storage change rate of the qth land use type in the region, in t / (hm 2 yr); x q Represents the total area of ​​the qth land use type in the region, in hm 2 .

[0133] Spatial constraints and area constraints are used as constraints for optimizing path regulation. The spatial constraint is that ecological land within the region cannot be converted into other land types. In order to make the multi-objective optimization of land use conform to the development law and policy constraints, based on the national land space, combined with the land use changes in the study area over the past 20 years, population evolution trends and related mathematical models, the constraints of various land types are constructed. The area constraint is that the total area of ​​each type of land is constant to the area of ​​the study area.

[0134] The non-dominated sorting genetic algorithm II (NSGA-II) is used to solve the multi-objective optimization model. The algorithm flow is as follows: ① Initialize the population and perform non-dominated sorting and crowding calculations on it. ② Select the parent generation, perform crossover and mutation to generate offspring. ③ Merge the parent generation and offspring, and select them according to non-dominated sorting and crowding to form a new generation of population. ④ Repeat steps ② and ③ until the termination condition is met.

[0135] Through multi-scenario comparative analysis, multi-objective optimization and multi-criteria decision-making models are combined with scenario analysis methods to optimize the spatial layout of the territory with the goal of reducing emissions and increasing carbon sinks in terrestrial ecosystems. It is necessary to reduce the loss of carbon storage in terrestrial ecosystems and focus on improving their carbon sink flow. Ultimately, multiple optimizations of the land use quantity structure, spatial layout, and management strategies to enhance regional vegetation carbon storage are achieved, so as to scientifically regulate and formulate the spatial distribution pattern and management policies for regional vegetation to maintain and enhance carbon storage and carbon sink potential.

[0136] This embodiment can optimize the maximum light energy efficiency parameters of the CASA model to construct multiple scenarios under the background of future climate change and land use change, improve the accuracy and reliability of regional vegetation carbon storage in each scenario, and provide multiple reference bases for the formulation of policies to increase regional vegetation carbon storage.

[0137] And by constructing multiple scenarios under the background of future climate change and land use change, setting multi-objective optimization goals in terms of the stock and flux of maximum vegetation carbon storage, considering different decision-making behaviors and decision-making environments in the process of spatial optimization configuration, and the complex nonlinear correlation or causal relationship between optimization factors, establishing a perspective on the quantity and pattern of land use type distribution, starting from the quantity, space, and use of regional territorial space, propose more scientific and accurate territorial space pattern optimization paths, regulation and management strategies for improving regional vegetation carbon storage.

[0138] The above is only an embodiment of the present invention, and the common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation includes the following steps: Step 1, determine the scope of the study area, collect relevant multi-source data sets required for regional vegetation carbon storage simulation and multi-scenario simulation, and the multi-source data include regional vegetation carbon storage simulation data and multi-scenario simulation data; Step 2: preprocess the vegetation data, meteorological data, land use data and climate scenario data in the multi-source data and build a database; It is characterized by further comprising: Step 3, modify the maximum light utilization parameters in the CASA model applicable to different ecosystems / land uses in the study area; Step 4: Compare the planned data with multi-period land use data to construct multiple scenarios of future climate change and land use change, and form 12 regional vegetation carbon storage estimation scenarios; Step 5, based on the combined characteristic data of 12 regional vegetation carbon storage estimation scenarios, the CASA model is used to calculate the regional vegetation carbon storage VCS under each scenario; Step 6, based on the regional vegetation carbon storage VCS of 12 regional vegetation carbon storage estimation scenarios, the trend analysis method is used to estimate the change trend of the regional vegetation carbon storage VCS; Step 7: Set the maximum values ​​of vegetation carbon storage and flux as the optimization targets for optimization path regulation, match different development paths with optimization targets based on scenarios, and conduct multi-path optimization and regulation of vegetation carbon storage in terms of regional land use quantity structure, spatial layout, and management strategy; A multi-objective optimization model is constructed with the maximum values ​​of vegetation carbon storage and flux. In the multi-objective optimization model, 12 scenarios are set, and the functions are realized as 2. The multi-objective optimization model has a total of 24 decision variables. The multi-objective optimization model has 2 objective functions, which are the maximum vegetation carbon storage and flux target. and maximum vegetation carbon storage flux targets , then the multi-objective optimization model is expressed as: ； The vegetation carbon stock inventory target is expressed as: ； In the formula, q and m represent the qth land use type and the number of land use types in the region, respectively; Represents the carbon density of the qth land use type in the region, in t / hm 2 ; Represents the total area of ​​the qth land use type in the region, in hm 2 ; The vegetation carbon storage flux target is expressed as: ； Where: q and m represent the qth land use type and the number of land use types in the region respectively; represents the rate of change of vegetation carbon storage in the qth land use type in the region; Represents the total area of ​​the qth land use type in the region, in hm 2 ; Spatial constraints and area constraints are used as constraints for optimizing the path control. The spatial constraint is that ecological land within the region cannot be converted into other land types, and the area constraint is that the total area of ​​each type of land is constant to the area of ​​the study area. The non-dominated sorting genetic algorithm II is used to solve the multi-objective optimization model; Through multi-scenario comparative analysis, multi-objective optimization and multi-criteria decision-making models are combined with scenario analysis methods to optimize the national land space layout with the goal of reducing emissions and increasing sinks in terrestrial ecosystems.

2. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 1 is characterized in that: In step 1, the regional vegetation carbon storage simulation data includes vegetation, meteorological, and light energy utilization efficiency parameter data required for estimating vegetation net primary productivity using the CASA model, the vegetation data includes the normalized vegetation index on a monthly scale, the meteorological data includes precipitation, temperature, and solar radiation, and the light energy utilization efficiency parameters include land use / cover type and maximum light energy utilization efficiency parameter data; The multi-scenario simulation data include climate scenario data and multi-period land use data of the coupled model comparison plan, and the climate scenario data include monthly-scale normalized difference vegetation index, precipitation, temperature, and solar radiation scenario data under the SSP1-2.6 low forcing scenario, the SSP2-4.5 medium forcing scenario, and the SSP5-8.5 high forcing scenario.

3. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 2 is characterized in that: In the step 2, the preprocessing includes: by comparing the projection information and pixel size of multi-source raster data of vegetation data, meteorological data, and land use data, based on the minimum raster pixel, using a series of operations such as format conversion, clipping, projection conversion, and resampling for preprocessing; By establishing the classification standard of land use according to the land resources and their utilization attributes, the land use types in the study area are divided into 6 categories, including cultivated land, forest land, grassland, water area, construction land, and unused land for pre-processing; Through the downscaling method, the multi-scenario simulation data of the large-scale, low-resolution global climate model output information are converted into small-scale, high-resolution regional ground climate change information. The simulation values ​​preprocessed by the downscaling method are compared with the historical observation data. The root mean square error, correlation coefficient, relative deviation coefficient and Nash-Sutcliffe model efficiency coefficient between the simulation values ​​and the observation values ​​are used to examine the simulation accuracy of temperature and precipitation. Random points are selected from space for correlation test the reliability of the simulation results.

4. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 3 is characterized in that: In the step 4, climate scenarios are constructed, and typical future climate change scenarios of low forcing scenarios, medium forcing scenarios, and high forcing scenarios are obtained based on coupled model comparison plan data; Based on the land use type data, the Markov model and PLUS model are used to predict the future land use change scenarios. According to the regional land use conversion law and combined with the future development plan of the region, four scenarios are set: natural development scenario, economic development scenario, ecological protection scenario, and sustainable development scenario. Based on four scenarios, the shared socio-economic development pathway and the typical concentration pathway RCPs combination scenario were coupled to construct 12 future climate change scenarios considering ground conditions.

5. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 4 is characterized in that: In step 4, the 12 regional vegetation carbon stock estimation scenarios include a low forcing scenario superimposed on a natural development scenario, a low forcing scenario superimposed on an economic development scenario, a low forcing scenario superimposed on an ecological protection scenario, a low forcing scenario superimposed on a sustainable development scenario, a medium forcing scenario superimposed on a natural development scenario, a medium forcing scenario superimposed on an economic development scenario, a medium forcing scenario superimposed on an ecological protection scenario, a medium forcing scenario superimposed on a sustainable development scenario, a high forcing scenario superimposed on a natural development scenario, a high forcing scenario superimposed on an economic development scenario, a high forcing scenario superimposed on an ecological protection scenario, and a high forcing scenario superimposed on a sustainable development scenario.

6. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 5 is characterized in that: In step 5, based on NDVI, temperature, precipitation, solar radiation, and land use data under 12 regional vegetation carbon storage estimation scenarios, the net primary productivity NPP of vegetation at a monthly scale within a preset time period in the region simulated by the CASA model; The growing biomass density and carbon storage of regional vegetation were calculated using the CASA model and vegetation net primary productivity (NPP).

7. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 6 is characterized in that: In step 6, the Theil-Sen Median trend method and the Mann-Kendall test method are combined to identify the long-term series change trend characteristics of regional carbon storage within a set time period, and 5 change trend results are obtained: significant degradation, slight degradation, basically stable, slight recovery, and significant recovery.

8. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 7 is characterized in that: In step 6, the Theil-Sen Median trend method is used to analyze the changing trend of regional vegetation carbon storage under 12 scenarios. The Theil-Sen Median trend calculation formula is: ； in, is the median function, , Represent the data of the jth and ith time series respectively; when When it is greater than 0, it indicates that the factor is in an increasing trend, otherwise it is a decreasing trend.

9. The method for optimizing the path of regional vegetation carbon storage based on multi-scenario simulation according to claim 7 is characterized in that: In step 6, the Mann-Kendall test method performs a trend test to identify the long-term series change trend characteristics of regional carbon storage within a set time period. The detected statistic S is expressed as: ； In the formula, S is a normal distribution with a mean of 0 and a variance of ; Time series data is a sample of n independent random variables with the same distribution. For all i, j ≤ n, and when i ≠ j, the distributions of Xi and Xj are different; When n>10, the standard normal system variable calculation formula is: ； Given a confidence level of α, if , then the null hypothesis is not accepted and the confidence level is set to ,when When , accept the original hypothesis. When , the sequence has a significant upward trend. , the series has a significant downward trend.

Images