Desert region carbon reserve prediction method and system based on PLUS-InVEST model

Abstract

The invention provides a desert area carbon reserve prediction method and system based on a PLUS-I nVEST model, and relates to the technical field of carbon reserve estimation and prediction. The desert region carbon reserve prediction method based on the PLUS-I nVEST model comprises an acquisition module used for acquiring land utilization data and carbon density data with equal year intervals in multiple periods, and driving factor and limiting factor data corresponding to the land utilization data; according to the method, the land utilization types are systematically and completely classified, the carbon density corresponding to each land utilization type is optimized, the estimation precision of the carbon reserves is improved, potential changes of driving factors and model parameter setting are fully considered, the simulation precision of future land utilization changes is higher, and the simulation efficiency is improved. The method can provide a theoretical basis for ecological construction and sustainable development of a research area, and is of great significance to optimization of a land utilization pattern of the research area and reasonable ecological decision making.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon reserve estimation and prediction, and in particular to a method and system for predicting carbon reserves in desert areas based on a PLUS-InVEST model. Background Art

[0002] At present, there are two main methods for assessing carbon storage in desert areas: field survey method and model assessment method. The field survey method is relatively accurate in estimating carbon storage in a small area, but due to its small scope of application, cumbersome operation process, and the possibility of affecting the local ecological environment during data collection, the model assessment method is usually given priority when assessing carbon storage in a larger area. Among the many models used for assessment, the InVEST model has been widely used worldwide because of its characteristics of requiring less data, fast operation speed, and accurate assessment results. The InVEST model is an ecosystem service function assessment model developed by Stanford University and jointly supported by The Nature Conservancy and the World Wide Fund for Nature. At present, the Carbon module in the InVEST model is the most widely used in calculating carbon storage. It can simulate the distribution of carbon storage at different time and space scales with high accuracy.

[0003] The popular application of land use models provides an innovative way to predict the dynamic changes of future carbon storage. Combining traditional land use simulation models such as FLUS and CA-Markov to simulate and predict future carbon storage has also become the focus of academic attention. However, these models have limitations in revealing the deep-level internal changes of land use, and it is difficult to accurately simulate the complex and changeable characteristics of land use type patches. The PLUS model is a new land use prediction model jointly developed by the School of Geography and Information Engineering of China University of Geosciences (Wuhan) and the High Performance Spatial Computing Intelligence Laboratory (HPSCIL) of the National GIS Engineering Technology Research Center. Compared with traditional simulation and prediction models, the PLUS model can more comprehensively explore various land use incentives and accurately simulate the spatiotemporal changes of multiple types of land use patches. Therefore, technicians in this field provide a method and system for predicting carbon storage in desert areas based on the PLUS-InVEST model to solve the problems raised in the above background technology. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the deficiencies of the prior art, the present invention provides a method and system for predicting carbon reserves in desert areas based on the PLUS-InVEST model, which solves the problem of prediction accuracy of traditional prediction methods.

[0006] (II) Technical solution

[0007] To achieve the above objectives, the present invention is implemented by the following technical solutions: A method for predicting carbon reserves in desert areas based on the PLUS-InVEST model, comprising:

[0008] An acquisition module is used to acquire land use data and carbon density data at equal intervals for multiple periods of years, as well as driving factor and limiting factor data corresponding to the land use data, wherein the driving factors include temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density and GDP value, and the limiting factors include water areas and construction land;

[0009] A calculation module, for evaluating the carbon storage of the study area through the InVEST model according to the land use data and the carbon density data, and obtaining the total carbon storage change characteristics of the study area;

[0010] An application module, for applying the predicted values ​​of each driving factor in the future simulation year and the limiting factor to the PLUS model;

[0011] The prediction module is used to extract the land use change parts of two periods in multiple years, set the model parameters according to the conversion between various land use types, and run the PLUS model to simulate future land use changes. The PLUS model is used to explore the impact of the driving factors of the land use changes on the spatial pattern of land use;

[0012] The analysis module is used to combine the prediction and analysis results of carbon storage changes in the study area under three different scenarios in the future and select the development scenario that is most conducive to increasing carbon storage in the study area;

[0013] Obtain land use data, carbon density data and driving and limiting factor data related to PLUS model simulation prediction in the study area at different periods. The driving factor data include temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density and GDP value; according to the topography of the desert area and the actual situation of the study area, water areas and construction land are defined as limiting factors;

[0014] Using the Carbon Storage and Sequestration module in the InVEST model, the carbon density of the four major carbon pools corresponding to different land use types in the study area and the land area raster data corresponding to different land use types were input to estimate the carbon storage over the years and obtain the change characteristics of the total carbon storage in the study area.

[0015] The neighborhood weights were set according to the expansion area proportions of different land use types in the study area. According to the needs of local sustainable development in the study area, three scenarios were set: natural development scenario ND, ecological protection scenario EP, and cultivated land protection scenario CP. The transfer matrix under different scenarios was established. Combined with driving factors and limiting factors, the land use changes in the study area under different scenarios in 2030 were predicted, and the impact of driving factors on the spatial pattern of land use was explored through the PLUS model.

[0016] The prediction results are brought back into the InVEST model to predict and analyze the changes in carbon storage in the study area under three different scenarios in the future, and the development scenario that is most conducive to increasing carbon storage in the study area is selected.

[0017] Preferably, the acquisition of multiple periods of land use data with equal year intervals, carbon density data corresponding to different land use types, and driving factor and limiting factor data corresponding to the land use data includes: acquiring four periods of land use data of the study area with equal year intervals in units of ten years; obtaining the carbon density value of each land use type through correction by publicly published literature and related formulas related to the study area; collecting driving factors corresponding to the years of the four periods of land use data; wherein the difference between the latest year and the current year when obtaining the land use data is no more than four years; when obtaining the carbon density, the existing carbon density research results in the study area are used as the main reference, and the carbon density data in the relevant research results of the neighboring areas of the study area are used for supplementation, and the carbon density correction formula proposed by Alam is used for correction in combination with the relevant research results of Li Kerang, Xie Xianli and others; for the driving factors whose attribute value changes in the past twenty years are negligible, only one period of data close to the latest year is collected, and the corresponding limiting factors are based on the latest year.

[0018] Preferably, the carbon density correction formula is as follows:

[0019] C BP =6.7981*e 0.00541*MAP

[0020] C BT =28*MAT+398

[0021] C SP =3.3968*MAP+3996.1

[0022]

[0023] K B =K BP *K BT

[0024]

[0025] Among them, MAP is the average annual precipitation, MAT is the average annual temperature, C BP , C BT They represent the biomass carbon density obtained based on the annual average precipitation and annual average temperature, respectively. SP is the soil carbon density corrected according to the average annual precipitation; K BP , K BT are the precipitation factor and temperature factor correction coefficients of biomass carbon density, K B , K S They are biomass carbon density correction factor and soil carbon density correction factor, respectively.

[0026] Preferably, the calculation formula for the total carbon storage is as follows:

[0027] C total =C above +C below +C soil +C dead

[0028] Among them, C total is the total carbon storage in the desert area; C above is the carbon storage of aboveground biological carbon pool; C below is the carbon storage of underground biological carbon pool; C soil is the carbon storage of soil organic carbon pool; C dead is the carbon storage of dead organic carbon pool.

[0029] Preferably, the land use simulation prediction data including land use data, driving factor data and influencing factor data are unified into tif format, and the projection coordinate system is changed to be consistent, wherein the row and column numbers of the land use data are unified, and the classification number is changed to start from 1; based on the pre-processed land use simulation prediction data, the land use change situation under three different scenarios of the study area is predicted by the PLUS model, and the influence of the driving factors of the land use change situation on the land use spatial pattern is analyzed by the PLUS model, including:

[0030] Constructing a PLUS model, wherein the PLUS model includes a land expansion analysis strategy module and a cellular automation model module based on multiple types of random patch seeds;

[0031] The land use data is transformed through Data Processing, and the successfully transformed data will have a suffix of "_uc";

[0032] The land use quantity is predicted by Markov-chain method to calculate the demand of the land use type;

[0033] The land expansion analysis strategy module LEAS is used to extract the expansion of various types of land between two phases of land use changes, and the random forest algorithm is used to mine the factors of land use expansion and driving force one by one to obtain the development probability of various types of land.

[0034] Through the CA module CARS based on multi-class random patch seeds combined with random seed generation and threshold reduction mechanism, the automatic generation of patches is simulated dynamically in time and space under the constraint of development probability to predict the land use change in the study area.

[0035] A desert area carbon stock prediction system based on the PLUS-InVEST model comprises: a memory and a processor, wherein the memory stores a program or instruction that can be run on the processor, and the processor implements the steps of the desert area carbon stock prediction method as described in any one of claims 1 to 5 when executing the program or the instruction.

[0036] (III) Beneficial effects

[0037] The present invention provides a method and system for predicting carbon reserves in desert areas based on a PLUS-InVEST model.

[0038] It has the following beneficial effects:

[0039] 1. In the present invention, by constructing different land use scenarios, the natural development, ecological protection and farmland protection scenarios selected in the embodiment can quantitatively predict the spatial and temporal distribution patterns of land use by coupling the PLUS-InVEST model; this multi-scenario prediction capability helps to evaluate the impact of different policies and land use changes on carbon storage, and provide a scientific basis for policy formulation and land use structure adjustment.

[0040] 2. In the present invention, not only land use changes are taken into account, but also natural environmental factors and socio-economic factors are combined to comprehensively analyze the driving mechanisms affecting the spatial differentiation of carbon storage; this method can more accurately reflect the comprehensive impact of multiple factors on carbon storage changes and provide a more comprehensive analysis framework.

[0041] 3. In the present invention, the PLUS model has high accuracy in simulating the generation and evolution of land use patches, which is superior to other models such as FLUS and CLUE-S; the InVEST model is known for its low data requirements and fast running speed; by coupling these two models, the present invention can efficiently predict the future land use spatial characteristics and carbon stock estimation of the study area, providing theoretical support for carbon management and national land space planning.

[0042] 4. The present invention provides a data-driven framework to help decision makers and researchers make more scientific and reasonable decisions based on the latest data and analysis results. In addition, the present invention also provides researchers with a powerful tool for analyzing the impact of different land use scenarios on carbon storage and further deepening the understanding of the mechanism of carbon storage changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a flow chart of the method for estimating and predicting carbon storage in desert areas based on the PLUS-InVEST model;

[0044] Figure 2 It is a schematic diagram of the implementation process of the present invention;

[0045] Figure 3 This is the spatial distribution map of carbon storage in the study area from 1990 to 2020;

[0046] Figure 4 This is a schematic diagram of the contribution of driving factors to different land use types in the study area from 1990 to 2020;

[0047] Figure 5 This is the spatial distribution map of carbon storage in the study area under different scenarios in 2030;

[0048] Figure 6 This is the schematic diagram of the device for estimating and predicting carbon storage in desert areas based on the PLUS-InVEST model;

[0049] Figure 7 This is the schematic diagram of the carbon storage estimation and prediction system in desert areas based on the PLUS-InVEST model;

[0050] Figure 8 This is the data source acquisition table for the study area in this invention;

[0051] Fig. 9 This is the carbon density table of each land use type in the study area of ​​this invention;

[0052] Fig.10 This is the table of different land requirements under three scenarios in the study area in 2030 in this invention;

[0053] Fig.11 This is a table for setting the transfer matrix under three scenarios for the study area in 2030 in this invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] Embodiment 1:

[0056] like Figure 1-11 As shown, the embodiment of the present invention provides a method for predicting carbon reserves in desert areas based on the PLUS-InVEST model, including:

[0057] The acquisition module is used to obtain land use data and carbon density data at equal intervals for multiple periods, as well as driving factor and limiting factor data corresponding to the land use data. The driving factors include temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density and GDP value. The limiting factors include water areas and construction land.

[0058] The calculation module is used to evaluate the carbon storage in the study area through the InVEST model based on the land use data and carbon density data, and obtain the total carbon storage change characteristics of the study area;

[0059] The application module is used to apply the predicted values ​​and limiting factors of each driving factor in the future simulation years to the PLUS model;

[0060] The prediction module is used to extract the land use change parts of two periods in multiple years, set the model parameters according to the conversion between various land use types, and run the PLUS model to simulate future land use changes. The PLUS model is used to explore the impact of the driving factors of land use changes on the spatial pattern of land use;

[0061] The analysis module is used to combine the prediction and analysis results of carbon storage changes in the study area under three different scenarios in the future and select the development scenario that is most conducive to increasing carbon storage in the study area;

[0062] Obtain land use data, carbon density data and driving and limiting factor data related to PLUS model simulation prediction in the study area at different periods. The driving factor data include temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density and GDP value; according to the topography of the desert area and the actual situation of the study area, water areas and construction land are defined as limiting factors;

[0063] Using the Carbon Storage and Sequestration module in the InVEST model, the carbon density of the four major carbon pools corresponding to different land use types in the study area and the land area raster data corresponding to different land use types were input to estimate the carbon storage over the years and obtain the change characteristics of the total carbon storage in the study area.

[0064] The neighborhood weights were set according to the expansion area proportions of different land use types in the study area. According to the needs of local sustainable development in the study area, three scenarios were set: natural development scenario ND, ecological protection scenario EP, and cultivated land protection scenario CP. The transfer matrix under different scenarios was established. Combined with driving factors and limiting factors, the land use changes in the study area under different scenarios in 2030 were predicted, and the impact of driving factors on the spatial pattern of land use was explored through the PLUS model.

[0065] The prediction results are brought back into the InVEST model to predict and analyze the changes in carbon storage in the study area under three different scenarios in the future, and the development scenario that is most conducive to increasing carbon storage in the study area is selected.

[0066] Acquire land use data with equal year intervals, carbon density data corresponding to different land use types, and driving factor and limiting factor data corresponding to the land use data, including: obtain four periods of land use data of the study area with equal year intervals in units of ten years; obtain the carbon density value of each land use type through correction by publicly published literature and related formulas related to the study area; collect driving factors corresponding to the years of the four-period land use data; the difference between the latest year and the current year when obtaining land use data should not exceed four years; when obtaining carbon density, the existing carbon density research results in the study area are used as the main reference, and the carbon density data in the relevant research results of the neighboring areas of the study area are used for supplementation, and the carbon density correction formula proposed by Alam is used in combination with the relevant research results of Li Kerang, Xie Xianli and others for correction; for the driving factors whose attribute value changes in the past twenty years are negligible, only the data of the period close to the latest year are collected, and the corresponding limiting factors are based on the latest year.

[0067] The carbon density correction formula is as follows:

[0068] C BP =6.7981*e 0.00541*MAP

[0069] C BT =28*MAT+398

[0070] C SP =3.3968*MAP+3996.1

[0071]

[0072]

[0073] Among them, MAP is the average annual precipitation, MAT is the average annual temperature, C BP , C BT They represent the biomass carbon density obtained based on the annual average precipitation and annual average temperature, respectively. SP is the soil carbon density corrected according to the average annual precipitation; K BP , K BT are the precipitation factor and temperature factor correction coefficients of biomass carbon density, K B , K S are the biomass carbon density correction factor and soil carbon density correction factor, respectively;

[0074] The four major carbon pools include aboveground biomass, underground biomass, dead organic matter and soil organic matter; the carbon density of aboveground biomass is the average carbon content of a fixed area within the range of 0 to 20 cm on the earth's surface; the carbon density of underground biomass is the average organic carbon content in the plant roots of a fixed area below the surface; the carbon density of soil organic matter is the organic carbon content per unit area of ​​soil 20 to 100 cm below the earth's surface. The calculation formula for total carbon storage is as follows:

[0075] C total =C above +C below +C soil +C dead

[0076] Among them, C total is the total carbon storage in the desert area; C above is the carbon storage of aboveground biological carbon pool; C below is the carbon storage of underground biological carbon pool; C soil is the carbon storage of soil organic carbon pool; C dead is the carbon storage of dead organic carbon pool.

[0077] The PLUS model was used to predict and analyze the land use changes in the study area under three different scenarios in the future, and the impact of driving factors on the spatial pattern of land use was explored through the PLUS model.

[0078] Specifically, according to the needs of local sustainable development in the study area, three scenarios, namely, natural development scenario ND, ecological protection scenario EP and cultivated land protection scenario CP, were set, and transfer matrices under different scenarios were established.

[0079] (1) Scenario 1: Natural development scenario ND; This scenario is based on the land use change from 2010 to 2020 and the Markov-Chain prediction results of the PLUS model. It predicts the land use demand in the study area without natural intervention in 2030 and uses it as the basis for simulating other scenarios;

[0080] (2) Scenario 2: Ecological protection scenario EP; This scenario takes ecological security and the maintenance of sustainable development of the ecosystem as its main purpose, and strictly protects grasslands, cultivated land and other areas that have a significant impact on the ecological environment; This paper combines the actual situation of the study area, and modifies the Markov transfer probability matrix on the basis of the natural development scenario, reducing the proportion of cultivated land and grassland converted to construction land. At the same time, considering that the ecological capacity of cultivated land is weaker than that of grassland, the proportion of cultivated land converted to grassland is increased, thereby achieving ecological protection;

[0081] (3) Scenario 3: Cultivated land protection scenario CP; This scenario aims to protect cultivated land area and ensure food security and ecosystem stability; based on the natural development scenario, the present invention protects cultivated land area by reducing the proportion of cultivated land converted to other land types and increasing the proportion of unused land converted to cultivated land, thereby achieving cultivated land protection;

[0082] The present invention takes the land use data of 2010 as the benchmark data, predicts the land use development in the study area in 2020, and uses the verification module of the PLUS model to verify the accuracy of the predicted results with the actual land use data in 2020, so as to ensure the accuracy of the simulation results of the PLUS model; according to the accuracy verification results, the overall accuracy is 0.935, and the FoM value is 0.271; according to the comparative analysis of previous research results, the FoM value of the existing land use change model is between 0.1-0.3, and the larger the FoM value, the higher the model accuracy; this result shows that the simulation accuracy and accuracy of the PLUS model in the study area are high, and it can be used to predict the results of land use changes in the study area under different scenarios in 2030.

[0083] Construct the PLUS model, which includes a land expansion analysis strategy module and a cellular automation model module based on multiple types of random patch seeds. The main operation process has the following steps:

[0084] (1) Data processing: In the previous operation of reclassifying LULC data using ArcGIS software, ArcGIS usually converts the original "unsigned char" LULC data into "int" or "unsigned int" format; therefore, it is necessary to convert the four-period land use data from 1990 to 2020 into "unsigned char" format through "Convert LULCs to Unsigned Char Format". The successfully converted data will have a "_uc" ​​suffix;

[0085] (2) Extract land expansion: Use “Extract Land Expansion” to extract the land expansion between 2010 and 2020, so as to prepare for the subsequent land expansion analysis strategy module LEAS;

[0086] (3) Land Expansion Analysis Strategy LEAS module; through the "Land Expansion Analysis Strategy LEAS" module, the land expansion data from 2010 to 2020 preprocessed in the previous step and the nine driving factor data selected by the present invention, namely, temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density and GDP value, are input, and the parameters of Random Forest Classification (RFC) are set at the same time; the sampling rate is set to 0.01, which means that about 1% of the pixels are selected for training; the number of regression trees is set to 20; the number of features mTry used to train the RFC model is set to 16. According to the five land use types divided in the study area in the present invention, the PLUS model outputs 5 development potential maps and the training accuracy of RFC for each land use;

[0087] (4) CA module CARS based on multi-class random patch seeds; input the land use data of 2020 as the starting year, the development potential data output in the previous step, and the two restriction factor data of water area and construction land set by the present invention, and set the following parameters:

[0088] ① Land demand; The land demand table simulates the number of grids of different land use types in the year, mainly including two simulation methods: linear regression method and Markov chain; the present invention uses the Markov chain method to calculate the number of grids of different land use types under three scenarios in 2030, and the calculation results are shown in Table 3;

[0089] ② Neighborhood weight; Neighborhood weight can reflect the expansion capacity of different land use types, and its value is between 0 and 1. The higher the value, the stronger the expansion capacity of the land use type. This paper sets the neighborhood weight according to the expansion area ratio of different land use types in the study area. It is calculated that the neighborhood weights of cultivated land, grassland, water area, unused land, and construction land are 0.243, 0.491, 0.006, 0.257, and 0.002, respectively.

[0090] ③ Transfer matrix; The purpose of setting the transfer matrix is ​​to quantify the conversion probability between different land use types, so as to simulate and predict the trend and pattern of land use change and provide a scientific basis for land planning and management; the specific setting rule is that when a land use type is allowed to transform into another land use type, its value in the transfer matrix is ​​set to 1, otherwise it is set to 0; the land use transfer matrix under the three scenarios set by the present invention is shown in Table 4;

[0091] After inputting all the above data, the automatic generation of patches is simulated dynamically in time and space under the constraint of development probability, and the land use changes under three different scenarios in 2030 are output; compared with the base year of 2020, the overall change trend under the natural development scenario in 2030 is consistent with the change trend of the study area from 1990 to 2020; the grassland area under the ecological protection scenario increases by 5.20km2 compared with the natural development scenario, an increase of 0.61%; the cultivated land and unused land decrease by 5.20km2 and 0.20km2 respectively compared with the natural development scenario, with a decrease of 0.59% and 0.02% respectively, which may be due to In this scenario, ecological protection is taken as the core goal, and the proportion of each type of land converted to grassland is set to increase; the core purpose of the farmland protection scenario is to ensure the safety of agricultural production while curbing the growth of non-agricultural land; under the farmland protection scenario, the cultivated land area increased by 10.25km2 compared with the natural development scenario, an increase of 1.16%, which shows that the farmland protection has achieved remarkable results; the unused land and grassland areas have changed slightly compared with the natural development scenario, among which the unused land area increased by 1.11km2, an increase of 0.12%; the grassland area decreased by 11.39km2, a decrease of 1.28%.

[0092] In addition, based on the RFC training accuracy of different land use types output by the LEAS module, the present invention evaluates and analyzes the contribution of 9 driving factors, including temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density and GDP value, to the expansion of different land use types in the study area from 1990 to 2020, and explores the key factors affecting the change of land use spatial pattern. The results are as follows: Figure 4As shown in the figure; the top three driving factors that contribute most to the expansion of cultivated land in the study area are DEM, distance from third-level roads, and distance from second-level roads; analyzing the reasons, firstly, cultivated land generally chooses areas with low altitude and flat terrain to grow crops, so it is relatively difficult to expand cultivated land in areas with high altitude and complex terrain; in addition, the impact of road proximity on cultivated land is also very important. Areas close to roads may be more vulnerable to the impact of urbanization and industrialization, resulting in the conversion of cultivated land to non-agricultural use. The main driving factor affecting grassland is DEM; the altitude is usually closely related to the type and fertility of soil. At higher altitudes, the soil is often poorer, which may limit the expansion of grassland; the main driving factors affecting unused land are GDP and DEM; first, a higher GDP often indicates a higher economic level in the region, which may increase the overall demand for land resources and promote the development and utilization of unused land; second, unused land located at high altitudes or with complex terrain is usually concentrated in the study area. Protected areas, which will also limit the development of these areas; DEM and population density are the main driving factors affecting construction land; usually with the increase in population density, in order to meet people's daily needs, the demand for urban construction space also increases; on the other hand, urban expansion tends to avoid areas with high altitudes and complex terrain, and prefer areas with less terrain undulations and lower development difficulty.

[0093] 4. Substituting the simulation results of the PLUS model into the InVEST model, the carbon storage changes in the study area under three different scenarios in 2030 were predicted and analyzed. The results are as follows: Figure 5 As shown; According to the prediction results, the development scenario that is most conducive to the increase of carbon storage in the study area is selected; Compared with the base year of 2020, carbon storage has increased under the three different scenarios; The overall change trend of carbon storage in various types of land under the natural development scenario is consistent with the change trend in the study area over the past 30 years; The carbon storage in the ecological protection scenario increases the most compared with 2020. The reason is that in this scenario, while restricting the transfer of cultivated land and grassland area, it increases the proportion of conversion of various types of land to grassland, which is more conducive to the increase of carbon storage in the study area; The carbon storage in the cultivated land protection scenario increases the least compared with 2020. It may be that although the expansion speed of cultivated land is increased in this scenario, the probability of conversion of other land types to grassland is also reduced to some extent. The carbon density of cultivated land is lower than that of grassland, which makes the carbon storage of grassland decrease instead of increase, and then leads to a relatively small increase in carbon storage; The results show that the ecological protection scenario is most conducive to the increase of carbon storage in the Haba Lake Nature Reserve.

[0094] like Figure 6 As shown, in an embodiment of the present invention, a carbon storage estimation and prediction device for desert areas based on the PLUS-I nVEST model includes an acquisition module 201, a calculation module 202, a prediction module 203 and an analysis module 204;

[0095] (1) An acquisition module 201 is used to acquire land use data and carbon density data at equal intervals for multiple periods of years, as well as driving factor and limiting factor data corresponding to the land use data. The driving factors include temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density, and GDP value. The limiting factors include water areas and construction land.

[0096] (2) prediction module 202, used to evaluate the carbon storage of the study area through the InVEST model based on the land use data and carbon density data, and obtain the total carbon storage change characteristics of the study area;

[0097] (3) Prediction module 203, used to extract the land use change parts of two periods in multiple years, set model parameters according to the conversion between various land use types, and run the PLUS model to simulate future land use changes, and explore the impact of the driving factors of land use change on the land use spatial pattern through the PLUS model;

[0098] (4) Analysis module 204 is used to combine the prediction analysis results of carbon storage changes in the study area under three different scenarios in the future and select the development scenario that is most conducive to increasing carbon storage in the study area.

[0099] The land use simulation prediction data including land use data, driving factor data and influencing factor data are unified into tif format, and the projection coordinate system is changed to be consistent. The row and column numbers of the land use data are unified, and the classification number is changed to start from 1. Based on the pre-processed land use simulation prediction data, the land use change in the study area under three different scenarios is predicted by the PLUS model, and the impact of the driving factors of land use change on the spatial pattern of land use is analyzed by the PLUS model, including:

[0100] Construct the PLUS model, which includes a land expansion analysis strategy module and a cellular automation model module based on multiple types of random patch seeds;

[0101] The land use data is transformed through Data Processing, and the successfully transformed data will have a suffix of "_uc";

[0102] The Markov-chain method is used to predict the quantity of land use and calculate the demand for land use types;

[0103] The land expansion analysis strategy module LEAS is used to extract the expansion of various types of land between two phases of land use changes, and the random forest algorithm is used to mine the factors of land use expansion and driving force one by one to obtain the development probability of various types of land.

[0104] Through the CA module CARS based on multi-class random patch seeds combined with random seed generation and threshold reduction mechanism, the automatic generation of patches is simulated dynamically in time and space under the constraint of development probability, and the land use change in the study area is predicted.

[0105] The simulation results of the PLUS model were substituted into the InVEST model to predict and analyze the changes in carbon storage in the study area under three different scenarios in the future, and the development scenario that is most conducive to increasing carbon storage in the study area was selected.

[0106] The desert area carbon storage prediction system based on the PLUS-InVEST model comprises: a memory and a processor, wherein the memory stores programs or instructions that can be run on the processor, and when the processor executes the programs or instructions, the steps of the desert area carbon storage prediction method as claimed in any one of claims 1 to 5 are implemented.

[0107] The present invention is further described below by taking Haba Lake National Nature Reserve as an example;

[0108] As an important ecological barrier in the agricultural and pastoral transition zone in eastern Ningxia, the Haba Lake Nature Reserve plays a vital role in preventing and controlling land desertification and maintaining ecological balance. The optimization of its ecological conditions has a far-reaching impact on the ecological security and development strategy of Ningxia and even the whole country. At the current critical juncture of promoting the realization of the "carbon peak" and "carbon neutrality" goals, exploring the carbon reserves in the Haba Lake Nature Reserve will provide a solid scientific basis for the local planning, supervision and protection work of the Haba Lake Nature Reserve.

[0109] like Figure 2 As shown in the figure, the method for estimating and predicting carbon storage in desert areas based on the PLUS-InVEST model mainly includes the following steps:

[0110] 1. Obtain the land use data of Habahu National Nature Reserve in 1990, 2000, 2010 and 2020, and use ArcGIS10.8 to unify the coordinate system, clip, unify the resolution to 30m, reclassify and unify the row and column numbers of the land use data to 1-5, and change the classification number to start from 1 and other preprocessing operations; According to the topography and landforms of the desert area and the actual situation of the study area, the land use in the study area is classified into five land use types: cultivated land, grassland, water area, unused land, and construction land;

[0111] The driving factors corresponding to the years of the four-period land use data were collected. Among them, the driving factors with negligible changes in attribute values ​​in the past twenty years only collected data close to the latest year. The corresponding limiting factors were based on the latest year. In combination with the actual situation of the study area and the availability of data, the present invention combined natural environmental factors and socio-economic factors, and selected 9 data items including temperature, precipitation, elevation, slope, distance to roads, distance to secondary and tertiary roads, distance to railways, population density and GDP value as the driving factors of land use change in the study area. The projection coordinate system of the driving factor data was changed to be consistent, and the pre-processing operations such as resolution, pixel size of the driving factor, number of rows and columns, etc. were unified. At the same time, water areas and construction land were defined as limiting factors to limit their conversion to other land types in the PLUS model. The sources of land use data and driving factor data are shown in Table 1.

[0112] The carbon density data of different land use types in the study area were collected. The present invention takes the existing research results of Ningxia as the main reference, and uses the carbon density data in the relevant research results of the adjacent areas of the study area to supplement, so as to ensure that the carbon density data required by the model is as accurate and comprehensive as possible; finally, the carbon density correction formula proposed by Alam et al. was used, combined with the relevant research results of Li Kerang, Xie Xianli et al., to obtain the carbon density values ​​of different land use types in the study area as shown in Table 2;

[0113] Repeat the above steps to estimate carbon storage over the years, and import the carbon density data obtained after running the InVEST model into ArcGIS. Through the zoning statistics function in ArcToolbox, the carbon storage change characteristics of different land use types in the study area from 1990 to 2020 are calculated as follows: Figure 3As shown in the figure, the carbon storage in the study area from 1990 to 2020 showed an overall upward trend. The carbon storage increased the least from 1990 to 2000, the most from 2000 to 2010, and the carbon storage increased in the middle from 2010 to 2020. Based on the actual situation of the study area, it may be because the construction of the study area began in 2006. Before that, the area had been in a state of natural non-intervention, which led to a relatively slow growth rate of carbon storage from 2000 to 2010. However, with the development of the study area in 2006, the growth rate of carbon storage from 2000 to 2010 has increased significantly. This change clearly shows that the establishment of the study area It has played an important role in promoting the growth of carbon storage, and its results in protecting the local environment are obvious; from the perspective of land use type, the total carbon storage is ranked from large to small as follows: grassland > cultivated land > unused land > construction land > water area, among which grassland has the largest carbon storage proportion and is the main carbon pool in the study area; followed by cultivated land carbon storage; grassland carbon storage has generally shown an upward trend over the past 30 years, and increased the most from 2000 to 2010; the carbon storage of cultivated land and unused land has generally shown a downward trend; construction land and water areas have a smaller area proportion and relatively small carbon density, so the overall carbon storage has not changed much.

[0114] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting carbon storage in desert areas based on the PLUS-InVEST model, characterized by: include: An acquisition module is used to acquire land use data and carbon density data at equal intervals for multiple periods of years, as well as driving factor and limiting factor data corresponding to the land use data. The driving factors include temperature, precipitation, elevation, slope, distance to roads (secondary and tertiary roads), distance to railways, population density and GDP value. The limiting factors include water areas and construction land. A calculation module, for evaluating the carbon storage of the study area through the InVEST model according to the land use data and the carbon density data, and obtaining the total carbon storage change characteristics of the study area; An application module, for applying the predicted values ​​of each driving factor in the future simulation year and the limiting factor to the PLUS model; The prediction module is used to extract the land use change parts of two periods in multiple years, set the model parameters according to the conversion between various land use types, and run the PLUS model to simulate future land use changes. The PLUS model is used to explore the impact of the driving factors of the land use changes on the spatial pattern of land use; The analysis module is used to combine the prediction and analysis results of carbon storage changes in the study area under three different scenarios in the future and select the development scenario that is most conducive to increasing carbon storage in the study area; Obtain land use data, carbon density data and driving and limiting factor data related to PLUS model simulation prediction in the study area at different periods. The driving factor data include temperature, precipitation, elevation, slope, distance to roads (secondary and tertiary roads), distance to railways, population density and GDP value; according to the topography of the desert area and the actual situation of the study area, water areas and construction land are defined as limiting factors; Using the Carbon Storage and Sequestration module in the InVEST model, the carbon density of the four major carbon pools corresponding to different land use types in the study area and the land area raster data corresponding to different land use types were input to estimate the carbon storage over the years and obtain the change characteristics of the total carbon storage in the study area. The neighborhood weights were set according to the expansion area proportions of different land use types in the study area. According to the needs of local sustainable development in the study area, three scenarios were set: natural development scenario (ND), ecological protection scenario (EP) and farmland protection scenario (CP). The transfer matrix under different scenarios was established. Combined with driving factors and limiting factors, the land use changes in the study area under different scenarios in 2030 were predicted, and the impact of driving factors on the spatial pattern of land use was explored through the PLUS model. The prediction results are brought back into the InVEST model to predict and analyze the changes in carbon storage in the study area under three different scenarios in the future, and the development scenario that is most conducive to increasing carbon storage in the study area is selected.

2. The method for predicting carbon reserves in desert areas based on the PLUS-InVEST model according to claim 1, characterized in that: The method of obtaining land use data with equal year intervals for multiple periods, carbon density data corresponding to different land use types, and driving factor and limiting factor data corresponding to the land use data includes: obtaining four periods of land use data of the study area with equal year intervals in units of ten years; obtaining the carbon density value of each land use type through correction by publicly published literature and related formulas related to the study area; collecting driving factors corresponding to the years of the four periods of land use data; wherein the difference between the latest year and the current year when obtaining the land use data is no more than four years; when obtaining the carbon density, the existing carbon density research results in the study area are used as the main reference, and the carbon density data in the relevant research results of the adjacent areas of the study area are used for supplementation, and the carbon density correction formula proposed by Alam is used in combination with the relevant research results of Li Kerang, Xie Xianli and others for correction; for the driving factors whose attribute value changes in the past twenty years are negligible, only one period of data close to the latest year is collected, and the corresponding limiting factors are based on the latest year.

3. The method for predicting carbon reserves in desert areas based on the PLUS-InVEST model according to claim 2, characterized in that: The carbon density correction formula is as follows: C BP =6.7981*e 0.00541*MAP C BT =28*MAT+398 C SP =3.3968*MAP+3996.1 K B =K BP *K BT Among them, MAP is the average annual precipitation, MAT is the average annual temperature, C BP , C BT They represent the biomass carbon density obtained based on the annual average precipitation and annual average temperature, respectively. SP is the soil carbon density corrected according to the average annual precipitation; K BP , K BT are the precipitation factor and temperature factor correction coefficients of biomass carbon density, K B , K S They are biomass carbon density correction factor and soil carbon density correction factor, respectively.

4. The method for predicting carbon reserves in desert areas based on the PLUS-InVEST model according to claim 2, characterized in that: The calculation formula for the total carbon storage is as follows: C total =C above +C below +C soil +C deAd Among them, C ToTal is the total carbon storage in the desert area; C above is the carbon storage of aboveground biological carbon pool; C below is the carbon storage of underground biological carbon pool; C soil is the carbon storage of soil organic carbon pool; C dead is the carbon storage of dead organic carbon pool.

5. The method for predicting carbon reserves in desert areas based on the PLUS-InVEST model according to claim 1, characterized in that: The land use simulation prediction data including land use data, driving factor data and influencing factor data are unified into tif format, and the projection coordinate system is changed to be consistent, wherein the row and column numbers of the land use data are unified, and the classification number is changed to start from 1; based on the pre-processed land use simulation prediction data, the land use change in the study area under three different scenarios is predicted by the PLUS model, and the impact of the driving factors of the land use change on the land use spatial pattern is analyzed by the PLUS model, including: Constructing a PLUS model, wherein the PLUS model includes a land expansion analysis strategy module and a cellular automation model module based on multiple types of random patch seeds; The land use data is transformed through Data Processing, and the successfully transformed data will have a "_uc" ​​suffix; The land use quantity is predicted by Markov-chain method to calculate the demand of the land use type; The land expansion analysis strategy module (LEAS) is used to extract the expansion of various types of land between two phases of land use changes, and the random forest algorithm is used to mine the factors of land use expansion and driving force one by one to obtain the development probability of various types of land. Through the CA module based on multi-class random patch seeds (CARS) combined with random seed generation and threshold reduction mechanism, the automatic generation of patches is simulated dynamically in time and space under the constraint of development probability to predict the land use change in the study area.

6. The desert area carbon storage prediction system based on the PLUS-InVEST model is characterized by: include: A memory and a processor, wherein the memory stores a program or instruction that can be run on the processor, and when the processor executes the program or the instruction, the steps of the method for predicting carbon reserves in desert areas as described in any one of claims 1 to 5 are implemented.

Images