A method for estimating carbon budget of forest cover change

Abstract

The application discloses a forest cover change carbon budget estimation method and belongs to the technical field of regional carbon emission of forest cover change. The application obtains medium-high resolution vegetation carbon density spatial distribution data of a target area through a CASA model and the principle that net primary productivity is in a positive linear relationship with forest biomass, combines the CASA and Bookkeeping models, realizes carbon flux estimation distribution mapping of forest cover change with spatial details at a regional scale based on vegetation and soil carbon density spatial distribution data, and simulates more accurate regional carbon budget influence results of forest cover change. The application introduces carbon data with spatial details in the CASA model into the Bookkeeping model to obtain carbon flux data with spatial details caused by forest cover change, solves the existing model limitation problem, and can scientifically adjust the deficiencies of forest management at a regional scale.

Description

Technical Field

[0001] This invention belongs to the technical field of regional carbon emissions from forest cover change, and specifically relates to a method for estimating carbon budget from forest cover change. Background Technology

[0002] The impact of forest cover change on carbon budget has received widespread attention. Existing technologies have explored the effects of forest cover change on forest carbon storage, but these studies are mostly conducted at global, national, or provincial scales, resulting in high uncertainty in carbon emission estimates. Therefore, reducing this uncertainty is a topic of great interest to the scientific community, particularly in quantifying the carbon budget caused by forest cover change at local and regional scales. Forest cover change affects the carbon cycle of forest ecosystems by altering their structure and function. Simultaneously, forest cover change significantly alters the material production methods of various parts of the ecosystem, collectively influencing carbon storage and release. Regarding the types of forest cover change, it can be mainly divided into two categories: afforestation and deforestation. Deforestation directly reduces forest carbon storage through activities such as deforestation, while afforestation increases carbon sinks to some extent. Furthermore, the impact of forest cover change on forest carbon storage can vary drastically depending on the area of ​​afforestation or deforestation. Therefore, accurately quantifying the impact of regional-scale forest cover change on regional forest carbon storage is necessary.

[0003] Methods for estimating forest carbon budget caused by forest cover change include inventory methods, eddy covariance methods, and ecosystem carbon cycle models. Inventory methods rely on vegetation biomass and soil carbon storage data from field surveys to assess the ecosystem carbon pool. However, this method is limited by sample density, data representativeness, and the lack of continuous long-term records, making it difficult to accurately reflect spatiotemporal changes. Eddy covariance (EC) is a method that uses meteorological observation techniques to determine CO2 flux at a specific altitude by calculating the covariance between physical quantity fluctuations and vertical wind speed fluctuations, thereby assessing ecosystem productivity. However, the observation range of eddy covariance flux is limited, typically less than 1 kilometer, leading to insufficient accuracy in regional-scale carbon budget estimation. Furthermore, the high cost of flux tower construction limits the frequency of observation station deployment. Flux towers are usually located in typical areas with good vegetation and flat terrain, which introduces uncertainty in extending spatially continuous carbon distribution maps on heterogeneous surfaces and complex terrains. Ecosystem carbon cycle models are classified into three categories according to their modeling methods: empirical statistical models, process models, and remote sensing models (light energy use efficiency models). Empirical models establish statistical relationships between variables related to vegetation growth (such as light and soil) and carbon cycle variables (such as productivity), and then extrapolate to regional scales using spatialized data (such as topography, soil, and meteorological maps) to estimate the spatial distribution of the carbon cycle, such as the Miami model developed in 1972. However, due to the complex nonlinear relationships among various environmental factors, it is difficult to establish a comprehensive empirical model. Process models treat the soil-plant-atmosphere continuum as a whole, simulating key ecological processes such as photosynthesis, respiration, evapotranspiration, water and nutrient cycling, as well as environmental disturbances, to estimate the productivity of terrestrial ecosystems. By combining meteorological and environmental parameters, models can predict terrestrial ecosystem productivity on a larger regional scale and assess the impact of future climate change on the carbon budget of terrestrial ecosystems. Although process models can reveal the mechanisms of vegetation productivity in depth, their complexity and the difficulty in obtaining the required parameters limit their application. Light use efficiency (LUE) models calculate the proportion of photoactive radiation using vegetation indices or leaf area indices retrieved from remote sensing, and then multiply it by the incident photosynthetically active radiation, maximum light use efficiency, and environmental limiting factors to obtain total primary productivity. Relatively mature light energy utilization efficiency (LUE) models include CASA, MODIS-GPP, EC-LUE, VPM, and 3-PGS. These models utilize a simple empirical relationship between vegetation indices and the proportion of solar effective radiation to achieve continuous spatial simulation by connecting process models and remote sensing data. However, these models typically have coarse spatial resolution and limited spatial representation of forest cover change, leading to significant discrepancies between simulated carbon storage and observed values.

[0004] Bookkeeping models are a method of tracking carbon fluxes generated by land-use transformation, deforestation, wildfires, and other land-use / land-cover change types using carbon accounting methods. Bookkeeping models provide an effective means of estimating carbon pools and fluxes resulting from land-use / land-cover change. However, bookkeeping models are typically used to calculate carbon estimates at the national and global scales. The smallest modeling unit is an ecoregion, making it difficult to use for estimating carbon budgets at regional scales. Furthermore, bookkeeping models apply the same carbon density to changes in the same land type, thus failing to present carbon flux data with spatial detail.

[0005] In conclusion, relying on a single model may not be able to fully capture the spatially detailed forest carbon dynamics at the regional scale. Summary of the Invention

[0006] In response to the problems existing in the prior art mentioned in the background section, this invention proposes a method for estimating carbon budget of forest cover change. By combining the Bookkeeping model with the CASA model, forest carbon density information with complete spatial details is obtained to further explore the impact of forest cover change on forest carbon storage.

[0007] Technical Solution: To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] A method for estimating carbon budget of forest cover change is proposed. By using the CASA model and the positive linear relationship between net primary productivity and forest biomass, medium- and high-resolution spatial distribution data of vegetation carbon density in the target area are obtained. Combining CASA and Bookkeeping models, based on the spatial distribution data of vegetation carbon density and soil carbon density, carbon flux estimation and distribution mapping of forest cover change with spatial details at the regional scale are realized to simulate more accurate regional carbon budget impact results of forest cover change.

[0009] The specific implementation steps are as follows:

[0010] S1: Collect and process Landsat and MODIS NDVI remote sensing data, land cover type data, monthly average temperature data, monthly total precipitation data, monthly total solar radiation data, and soil spatial distribution data;

[0011] S2: Net primary productivity of the target region is simulated based on the CASA model and land cover type data;

[0012] S3: Collect and process the average carbon density of each land type in the target area, and use the positive linear relationship between net primary productivity and forest biomass to calculate forest vegetation carbon density data;

[0013] S4: Collect average soil carbon density data for the target area and calculate the change in soil organic carbon storage in the forest cover change area by multiplying litter productivity and litter turnover rate.

[0014] S5: Estimate and analyze the carbon budget caused by forest cover change based on the improved Bookkeeping model.

[0015] As a preferred option, in S1, the specific content of collecting and processing Landsat and MODIS NDVI remote sensing data, land cover type data, monthly average temperature data, monthly total precipitation data, monthly total solar radiation data, and soil spatial distribution data is as follows:

[0016] The collected NDVI data were synthesized using a high-quality reconstruction algorithm that interpolates missing data from remote sensing time-series images to generate monthly-scale NDVI time-series data.

[0017] We acquired monthly precipitation datasets and monthly average temperature datasets with 1 km resolution from the National Tibetan Plateau Scientific Data Center in China, and then interpolated them using the Kriging method to ensure that each dataset has a spatial resolution of 30 meters. We also generated monthly total solar radiation data from the long-sequence, high-density, and high-precision daily average solar radiation dataset based on station estimates, and interpolated the monthly total solar radiation data using the inverse distance weighting method to ensure that the data has a spatial resolution of 30 meters.

[0018] As a preferred option, in S2, the specific details of the net primary productivity of the target area simulated based on the CASA model and land cover type data are as follows:

[0019] First, land cover type data are used to detect forest cover changes in the target area to obtain spatiotemporal distribution data of forest cover changes;

[0020] Then, the model parameters were adjusted based on the maximum light energy utilization parameter to adapt them to the forests in the target area. Land cover type data, monthly NDVI time series data, monthly average temperature, monthly total precipitation and monthly total solar radiation data were imported into the CASA model to obtain the spatial distribution data of net primary productivity in the target area over the years.

[0021] As a preferred option, in S3, the average carbon density of each land type in the target area is collected and processed, and the forest vegetation carbon density data is calculated using the positive linear relationship between net primary productivity and forest biomass. The specific details are as follows:

[0022] The formula for calculating forest vegetation carbon density is as follows:

[0023]

[0024] Where i represents a pixel; C t N represents the average carbon density of forest vegetation. i C represents the net primary productivity of cell i; i Represents the forest vegetation carbon density of pixel i; n t This represents the total number of pixels occupied by forests in the target area; ∑ i∈t N i This represents the total net primary productivity of forest pixels.

[0025] As a preferred option, in S4, the average carbon density data of soil types in the target area are collected, and the changes in soil organic carbon storage in the forest cover change area are calculated by multiplying litter productivity and litter turnover rate.

[0026] By analyzing the litter decomposition process and employing a fixed first-order decomposition rate, the changes in soil organic carbon storage in forest cover change areas are calculated by multiplying litter productivity and litter turnover rate. The formulas for calculating changes in litter productivity and soil organic carbon storage are as follows:

[0027]

[0028] ΔSOCD=ΔLitter n ×TR L ,

[0029] Where n represents the nth year of each cycle, ΔLitter n NPP represents the litter productivity in year n. As and NPP Bs These refer to net primary productivity above and below ground, respectively; TR A and TR B These represent the turnover rates of aboveground and belowground biomass, respectively; ΔSOCD represents the change in soil organic carbon storage, and TR... L The turnover rate of litter biomass.

[0030] As a preferred approach, the reliability of the forest vegetation carbon density simulated by the CASA model is verified using field data from the target area in previous years. The specific details are as follows:

[0031] To verify the accuracy of vegetation carbon density estimated from the CASA model, correlation statistical analysis was performed to compare the forest vegetation carbon density estimated by the CASA model in previous years with the measured forest vegetation carbon density data from the same year's small-plot survey. Sample points of forest plot data in the target area were selected by random sampling, and the aboveground biomass of the forest at the sample points was extracted and multiplied by the forest vegetation carbon content coefficient of 0.5 to convert it into forest vegetation carbon density data.

[0032] As a preferred option, in S5, the specific content of estimating and analyzing the carbon budget caused by forest cover change based on the improved Bookkeeping model is as follows:

[0033] The formula for calculating the carbon budget caused by forest cover change is as follows:

[0034]

[0035] Where M represents the carbon budget caused by forest cover change; m represents the total number of years studied; ΔSOC n ΔVC represents the change in soil carbon storage in year n. n This represents the change in vegetation carbon storage in year n.

[0036] The formula for calculating the change in soil carbon storage caused by forest cover change is as follows:

[0037]

[0038] Wherein, ΔSOC n Let A represent the change in soil carbon storage in year n, where c and j represent land cover type c and another land cover type j, respectively, and the conversion between the two represents the change in forest cover type; A c,j,n It is the area where land cover type c changes to land cover type j in year n; ΔSOCD c,j,n This represents the change in soil organic carbon density when land cover type c changes to land cover type j in year n.

[0039] Preferably, changes in vegetation carbon storage caused by land cover change involve two main processes: an increase in vegetation organic carbon through restoration or artificial planting after the removal of native vegetation; and the release of removed vegetation biomass into the atmosphere as carbon dioxide at different oxidation rates after being used for construction, furniture manufacturing, or as carbon fuel. The calculation formula is as follows:

[0040] ΔVC n =ΔVC RS -ΔVC RM ,

[0041]

[0042] Wherein, ΔVC n It represents the change in vegetation carbon storage when a certain land cover type transforms into another land cover type in year n; ΔVC RS and ΔVC RM These represent the changes in vegetation carbon storage in year n due to biomass restoration and removal, respectively; A c，j，n It is the area where land cover type c changes to land cover type j in year n; VCD cLet α be the rate of change of vegetation carbon density for land cover type c. k It is the carbon storage of vegetation and in x k The ratio of total carbon reserves at different oxidation rates; x k This represents the oxidation rate of the removed vegetation under k different conditions; k represents the oxidation state of the removed vegetation.

[0043] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0044] (1) This invention obtains historical medium-to-high resolution spatial distribution data (30 meters) of forest vegetation carbon density in the target area using the CASA model and the principle that net primary productivity and forest biomass have a positive linear relationship. The 30-meter resolution spatial data of forest vegetation carbon density provided by this invention has higher spatiotemporal resolution than previous methods that obtained plot-scale vegetation carbon density data or discontinuous spatial distribution data. It not only provides quantitative information on forest vegetation carbon density but also more comprehensively reflects the spatial details and temporal characteristics of forest vegetation carbon density distribution in the target area. Simultaneously, importing the historical 30-meter spatial resolution forest vegetation carbon density data of the target area into the vegetation carbon pool of the Bookkeeping model improves the accuracy and timeliness of the model's estimation of carbon budget dynamics of forest cover change. The medium-to-high resolution spatial distribution data (30 meters) of forest vegetation carbon density obtained by this invention effectively solves the problem that previous methods could not obtain vegetation carbon density data with high resolution and regional-scale spatial details.

[0045] (2) Unlike the Bookkeeping model, which uses static carbon density data to calculate carbon budget, the method of this invention combines CASA and the Bookkeeping model. Based on the spatial distribution data of vegetation and soil carbon density, it improves the estimation and mapping of carbon fluxes of forest cover change at the regional scale with spatial details, so as to simulate a more accurate regional carbon budget impact of forest cover change. Compared with the traditional Bookkeeping model, the advantages of this invention are mainly reflected in the following aspects:

[0046] 1. Improved Data Accuracy: The CASA model can synthesize historical net primary productivity (NPP) for a target area using remote sensing and meteorological data. By leveraging the positive linear relationship between NPP data and forest biomass, spatially detailed vegetation carbon density is obtained, thus more accurately reflecting the spatiotemporal variations of forest vegetation carbon density. Simultaneously, by calculating litter decomposition processes using NPP data, changes and spatial distribution of soil organic carbon storage in forest cover change areas can be obtained. Introducing carbon data from the CASA model can effectively improve the accuracy of the Bookkeeping model in estimating the carbon budget caused by forest cover change.

[0047] 2. Overcoming Model Limitations: Previous Bookkeeping models only emphasized the comparison of carbon density between different land use types and their static spatial variability, failing to obtain spatially detailed carbon flux data. This invention imports spatially detailed carbon data from the CASA model into the Bookkeeping model to obtain spatially detailed carbon flux data caused by forest cover change. This overcomes this limitation, generating spatially detailed carbon flux data, which allows for scientific adjustments to shortcomings in regional-scale forest management, providing strong technical support for achieving the "dual carbon" target. This significantly compensates for the limitations of the Bookkeeping model at the regional scale. Attached Figure Description

[0048] Figure 1 This is a technical flowchart of the present invention;

[0049] Figure 2 This is a spatial distribution map of forest cover change generated from CLCD data of the present invention, wherein (a) is a distribution map of forest cover change from 2000 to 2010; (b) is a distribution map of forest cover change in the target area from 2010 to 2022; and (c) is a distribution map of forest cover change in the target area from 2000 to 2022.

[0050] Figure 3 This is a spatial distribution map of net primary productivity estimated by the CASA model of the present invention, wherein A is the net primary productivity distribution map in 2000; B is the net primary productivity distribution map in 2010; and C is the net primary productivity distribution map in 2022.

[0051] Figure 4 This invention presents a statistical chart of the area changes of major land cover types in Shaoguan City, Guangdong Province.

[0052] Figure 5 This is a comparative analysis of the correlation between the measured vegetation carbon density in a small plot and the vegetation carbon density simulated by the CASA model in this invention.

[0053] Figure 6 This is a carbon flux distribution map caused by forest cover changes in Shaoguan City, Guangdong Province from 2000 to 2022, which is the subject of this invention.

[0054] Figure 7 This is a statistical chart of carbon budget caused by different forest cover change types in Shaoguan City, Guangdong Province from 2000 to 2022. Detailed Implementation

[0055] The present invention will be further illustrated below with reference to specific embodiments. These embodiments are implemented based on the technical solutions of the present invention, and it should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention.

[0056] like Figure 1 As shown, the forest cover change carbon budget estimation method provided in this embodiment includes the following steps:

[0057] S1: Collect and process Landsat and MODIS NDVI remote sensing data, land cover type data, monthly average temperature data, monthly total precipitation data, monthly total solar radiation data, and soil spatial distribution data. Specific content includes:

[0058] Taking the forest cover change in Shaoguan City, Guangdong Province as an example:

[0059] NDVI remote sensing data: NDVI (Normalized Difference Vegetation Index) remote sensing data from Landsat and Moderate-Resolution Imaging Spectroradiometer (MODIS) from 2000 to 2022 were acquired using the Google Earth Engine (GEE) platform. The above NDVI data were then synthesized using the High-Quality Reconstruction Algorithm (GF-SG) for interpolating missing data from remote sensing time-series images to generate monthly-scale NDVI time-series data (30 meters).

[0060] Land cover type data: The data comes from the annual China Land Cover Dataset (CLCD) for 2000, 2010, and 2022, published by Yang Jie of Wuhan University as "China 30-meter Annual Land Cover Dataset and Its Dynamic Changes (1985-2022)". This data is sourced from the National Data Center for Glacier, Permafrost and Desert Science (http: / / www.ncdc.ac.cn). The forest cover change area in Shaoguan from 2000 to 2022 was obtained and its area was statistically analyzed using the land use transfer matrix.

[0061] Table 1 CLCD Land Cover Data Classification System

[0062] CLCD type name code farmland 1 forest 2 shrub 3 grassland 4 water body 5 Icefield 6 bare land 7 impermeable layer 8 wetlands 9

[0063] Table 2. Area of ​​forest cover change

[0064] 2000-2010 Area / hectare 2010-2022 Area / hectare 2000-2022 Area / hectare Forest-farmland 43639 Forest-farmland 58410 Forest-farmland 102049 Forest-shrub 293.4 Forest-shrub 352.5 Forest-shrub 645.9 forest-grassland 144.3 forest-grassland 42.8 forest-grassland 187.1 Forest - Impermeable Layer 1259.1 Forest - Impermeable Layer 1532.4 Forest - Impermeable Layer 2791.5 Farmland-Forest 18223 Farmland-Forest 26618 Farmland-Forest 44841 Shrub-Forest 837 Shrub-Forest 535.2 Shrub-Forest 1372.2 grassland-forest 141.8 grassland-forest 265.5 grassland-forest 407.3 Impermeable layer - forest 1.1 Impermeable layer - forest 2 Impermeable layer - forest 3.1

[0065] Monthly average temperature data, monthly total precipitation data, and monthly total solar radiation data: Obtained monthly precipitation datasets (1901-2023) and monthly average temperature datasets (1901-2023) with 1 km resolution from the National Tibetan Plateau Scientific Data Center (https: / / data.tpdc.ac.cn), and generated monthly total solar radiation data from long-sequence, high-density, and high-precision daily average solar radiation datasets (1960-2021) estimated based on stations (missing data for 2022 were filled with solar radiation data from 2021).

[0066] Then, the Kriging method was used to interpolate the monthly precipitation dataset (1901-2023) and the monthly average temperature dataset (1901-2023) of China at 1 km resolution, respectively. The inverse distance weighting (IDW) method was used to interpolate the total monthly solar radiation data to ensure that each data has a spatial resolution of 30 meters.

[0067] Soil spatial distribution data: The soil spatial distribution data (1:1,000,000) from the Second National Soil Census of the Chinese Academy of Sciences Resources and Environment Science and Data Center (https: / / www.resdc.cn / ) was converted to a spatial resolution of 30 meters by vector-to-raster conversion.

[0068] S2: Spatial distribution data of net primary productivity (NPP) in the target area are obtained based on the CASA model and land cover type data;

[0069] First, the annual China Land Cover Dataset (CLCD) was used to detect forest cover change in the target area, obtaining spatiotemporal distribution data of forest cover change. Second, land cover type data (CLCD), monthly NDVI time series data (30 meters), monthly average temperature (30 meters), monthly total precipitation (30 meters), and monthly total solar radiation (30 meters) were combined and imported into the CASA model to obtain spatial distribution data of net primary productivity (NPP) in the target area (Shaoguan City, 2000-2022).

[0070] Based on the processed data, the model parameters were adjusted using the optimal light energy utilization rate proposed by Zhu Wenquan's team to best suit the forests of the target area. This data was then imported into the improved CASA model developed by Zhu Wenquan's team to obtain the spatial distribution data of net primary productivity (NPP) over the years in the target area (e.g., ...). Figure 3(As shown). The optimal light energy utilization rate adjustment model parameters and the improved CASA model are derived from the paper "Remote Sensing Estimation of Net Primary Productivity of Terrestrial Vegetation in China" published by Zhu Wenquan's team in the Acta Ecologica Sinica. The paper includes the data for the optimal light energy utilization rate model parameters and the improved CASA model.

[0071] Table 3 Parameters for the maximum light energy utilization rate of the CASA model

[0072] Land cover type Maximum light energy utilization farmland 0.415 forest 0.985 shrub 0.429 grassland 0.389 water body 0 bare land 0.389 impermeable layer 0.389

[0073] S3: Collect and process the average carbon density of each land cover type in the target area, and use the positive linear relationship between net primary productivity and forest biomass to calculate forest vegetation carbon density data;

[0074] Based on collected literature data, the average carbon density of each land cover type in the target area (Shaoguan) was processed, including soil carbon density and vegetation carbon density data. Among them, the vegetation carbon density data came from the paper entitled "Carbon Emissions from Land-Use Change and Management in China between 1990 and 2010" published by Professor Huang Xianjin of Nanjing University and the research paper entitled "Spatiotemporal Changes of Carbon Storage in the Yellow River Basin Based on InVEST and CA-Markov Models" published by Yang Jie.

[0075] Table 4 Average Vegetation Carbon Density for Each Land Cover Type

[0076] Land cover type Average vegetation carbon density tC / ha <![CDATA[Agriculture field > 10.6 <![CDATA[Sen Forest > 47.2 shrub 11.7 grassland 7.7 bare land 2.1 impermeable layer 2.5

[0077] Based on the principle proposed in Kindermann et al.'s paper "A Global Forest Growing Stock, Biomass and Carbon Map Based on FAO Statistics" that there is a positive linear relationship between forest NPP and forest biomass, the distribution of forest vegetation carbon density in Shaoguan was calculated using the average carbon density data of Shaoguan forests. This carbon density was then combined with the average carbon density of other land types to obtain the vegetation carbon density distribution map of Shaoguan.

[0078] The formula for calculating forest vegetation carbon density is as follows:

[0079]

[0080] Where i represents a pixel; C t N represents the average carbon density of forest vegetation.i C represents the net primary productivity (NPP) of cell i; i Represents the forest vegetation carbon density of pixel i; n t This represents the total number of pixels occupied by forests in the target area; ∑ i∈t N i This represents the total net primary productivity of forest pixels.

[0081] S4: Collect average carbon density data of soil types in the target area, and calculate the change in soil organic carbon storage in the forest cover change area by multiplying litter productivity and litter turnover rate.

[0082] Soil carbon density was calculated by obtaining average carbon density data (0-20 cm) for soil types in Shaoguan from literature sources and integrating it with spatial distribution data for the corresponding soil types. The literature data for soil carbon density came from a paper titled "Number of soil profiles needed to give areliable overall estimate of soil organic carbon storage using profile carbondensity data" co-authored by Yan Xiaoyuan of Nanjing Institute of Soil Science and Professor Cai Zucong of Nanjing University.

[0083] Table 5 Average soil carbon density for each soil type

[0084] Soil type Soil carbon density (0-20 cm) tC / hectare Dark brown soil 58.7 swamp soil 18.0 Alluvial soil 19.2 Yellow-brown soil 32.4 Purple soil 21.1 lime soil 51.3 Lateritic red soil 35.3 paddy soil 31.2 Yellow soil 39.5 Yellow-brown soil 16.9 <![CDATA[Red Soil > 18.6 volcanic ash 37.4 Coarse bone soil 19.9 Red soil 33.4

[0085] By analyzing the litter decomposition process and employing a fixed first-order decomposition rate, the changes in soil organic carbon storage in forest cover change areas are calculated by multiplying litter productivity and litter turnover rate. The formulas for calculating changes in litter productivity and soil organic carbon storage are as follows:

[0086]

[0087] ΔSOCD=ΔLitter n ×TR L ,

[0088] Where n represents the nth year of each cycle, ΔLitter n NPP represents the litter productivity in year n. As and NPP Bs These refer to net primary productivity (NPP) above and below ground, respectively. TR A and TR B These are the turnover rates of aboveground and belowground biomass, respectively. NPP of grasslands and impermeable layers.As With NPP Bs The ratio is estimated to be 1:3, NPP of forest and shrubland As With NPP Bs The ratio is estimated to be 4:1. The harvest coefficient for farmland is 0.3. Furthermore, the root turnover rate for all vegetation types is calculated to be 0.15 per year. For grassland, farmland, and impermeable layers, the aboveground biomass turnover rate is 1.0 per year, while for forest it is 0.10 per year. Here, ΔSOCD represents the change in soil organic carbon storage, and TR... L The turnover rate of litter biomass.

[0089] The reliability of the forest vegetation carbon density simulated by the CASA model was verified using data from the 2020 Shaoguan small-scale forest survey (i.e., the second-class forest resource survey).

[0090] The 2020 forest vegetation carbon density data for Shaoguan City were obtained from the Guangdong Provincial Forestry Survey and Planning Institute. To verify the accuracy of the vegetation carbon density estimated based on the CASA model, a correlation statistical analysis was conducted to compare the model-estimated 2020 forest vegetation carbon density in Shaoguan with the measured forest vegetation carbon density data from the same year's forest vegetation carbon content coefficient (sample points from the 2020 forest vegetation carbon content coefficient in Shaoguan were selected using random sampling, and the aboveground biomass at each sample point was extracted and multiplied by 0.5 to convert it into forest vegetation carbon density data). The results showed a significant correlation between the 2020 forest vegetation carbon density estimated by NPP obtained from the CASA model and the measured forest vegetation carbon density values ​​from the same year, with R... 2 The value is 0.64 (see Figure 5 This indicates that the forest vegetation carbon density estimated by NPP based on the CASA model has a good fit and high correlation with the measured forest vegetation carbon density value, confirming that the carbon density value estimated based on the CASA model is suitable for subsequent analysis.

[0091] S5: Estimating and analyzing the carbon budget caused by forest cover change based on the improved Bookkeeping model;

[0092] By combining soil and vegetation carbon density change data in forest cover change areas, we improved the driven bookkeeping model to obtain the carbon budget caused by forest cover change. We divided forest cover change into two types: afforestation and deforestation, and analyzed the carbon budget (carbon loss and carbon sequestration) caused by forest cover change, while comparing the impact of afforestation and deforestation on the carbon budget.

[0093] The carbon budget caused by forest cover change from 2000 to 2022 is estimated based on the Bookkeeping model, and the calculation formula is as follows:

[0094]

[0095] The bookkeeping model consists of two carbon pools: vegetation and soil. Here, M represents the carbon budget due to forest cover change; m represents the total number of years studied; and ΔSOC... n ΔVC represents the change in soil carbon storage in year n. n This represents the change in vegetation carbon storage in year n.

[0096] Table 6 Bookkeeping Model Parameters

[0097] Land type forest shrub grassland impermeable layer Carbon content of dead vegetation in the soil during removal 33 50 100 49 Proportion of different oxidation rates of vegetation after removal 1 year 40 40 51 10 years 20 10 100 years 7 The percentage (%) of soil organic carbon that increased rapidly after cleanup. 16 16 16 23.15 Number of years required for soil organic carbon to increase rapidly after cleanup 15 15 15 5 Maximum reduction in soil organic carbon after cleanup (%) 20 4 4 28.94 Number of years required to reduce soil organic carbon to its maximum value after cleanup 45 45 45 10

[0098] Table 7 Definitions of Forest Cover Change Types

[0099] Forest cover change types <![CDATA[Sen Before the change in forest cover > <![CDATA[Sen After the forest cover changed > afforestation Any land type other than forest <![CDATA[Sen Forest > Deforestation <![CDATA[Sen Forest > Any land type other than forest

[0100] The formula for calculating changes in soil carbon storage caused by forest cover change is as follows:

[0101]

[0102] Wherein, ΔSOC n This represents the change in soil carbon storage in year n; c and j represent land cover type c and another land cover type j, respectively, and the conversion between the two represents the change in forest cover; A c，j，n It is the area in year n where land cover type c transforms into another land cover type j; ΔSOCD c,j,n This represents the change in soil organic carbon density when land cover type c changes to another land cover type j in year n.

[0103] Changes in vegetation carbon storage caused by land cover change involve two main processes: first, the increase in vegetation organic carbon through restoration or artificial planting after the removal of native vegetation; and second, the removal of vegetation biomass, which may be used for construction, furniture manufacturing, or as carbon fuel, ultimately releasing into the atmosphere as carbon dioxide at varying oxidation rates. The calculation formula is as follows:

[0104] ΔVC n =ΔVC RS -ΔVC RM ,

[0105]

[0106] Wherein, ΔVC n It represents the change in vegetation carbon storage in year n when the land is converted to other land types; ΔVC RS and ΔVC RM These represent the changes in vegetation carbon storage in year n due to biomass restoration and removal, respectively; A c,j,nIt is the area in year n where land cover type c transforms into another land cover type j; VCD c Let α be the rate of change of vegetation carbon density for land type c. k It is the carbon storage of vegetation and in x k The ratio of total carbon reserves at different oxidation rates. k This represents the oxidation rate of the removed vegetation under k different conditions; k represents the oxidation form of the removed vegetation, including burning firewood (1 year), paper oxidation (10 years), and oxidation of building materials and furniture (100 years).

[0107] S5: Results and Analysis;

[0108] like Figure 2 As shown in (a) above, the spatiotemporal dynamics of forest cover change:

[0109] Using CLCD data, the spatial distribution of forest cover change in Shaoguan City, Guangdong Province from 2000 to 2022 was identified and the area was statistically analyzed through the land use transfer matrix.

[0110] The results showed that the total forest type in Shaoguan changed by 165,707.6 hectares, accounting for approximately 9% of the total area. During the period 2000-2022, the forest area in Shaoguan decreased by 59,049.9 hectares. In contrast, the areas of farmland and impermeable land increased by 46,946.6 hectares and 13,868.5 hectares, respectively. Analysis of the areas of forest cover change in Shaoguan showed that the forest area decreased during the period 2000-2022 (e.g., ...). Figure 4 (As shown). The loss of forest area in Shaoguan is mainly driven by farmland expansion, accounting for 96.6% of the total forest area loss. Analysis of the impermeable layer in Shaoguan shows an increase in its area, indicating a rise in urbanization. Simultaneously, the area of ​​afforestation in the Shaoguan area shows an increasing trend, a result confirmed by the Forestry Statistical Yearbook.

[0111] like Figure 2 As shown in (b), the carbon budget caused by forest cover change:

[0112] To verify the accuracy of the vegetation carbon density estimated by the CASA model, a correlation statistical analysis was conducted to compare the forest vegetation carbon density simulated by the CASA model with the forest vegetation carbon density data measured in a small-scale survey in Shaoguan City in 2020. The results showed a significant relationship between the vegetation carbon density estimated by the CASA model's NPP results and the carbon density values ​​from the small-scale survey, with R0... 2 It is 0.64 (e.g.) Figure 5 (As shown). This indicates that the vegetation carbon density estimated by the model fits well with the measured vegetation carbon density value and has a high correlation, confirming the reliability of the vegetation carbon density value obtained by the CASA model.

[0113] like Figure 6 , 7 As shown, the carbon budget caused by forest cover change was estimated based on the Bookkeeping model. The results indicate that from 2000 to 2022, Shaoguan City experienced a net carbon loss due to forest cover change, with forest cover change in Shaoguan resulting in a carbon loss of 1.293 TgC. The carbon flux caused by forest cover change in Shaoguan City showed significant spatial differences at the regional scale (see...). Figure 6 The carbon flux in Shaoguan is mainly distributed in the western, central, northeastern, and southern regions, with a relatively dense distribution. Specifically, the central and western regions mainly show carbon loss. The northeastern region shows carbon sequestration. In the southern region, both carbon loss and carbon sequestration occur. Carbon loss and carbon sequestration were monitored in both the northwestern and southern regions. To study the carbon flux caused by different types of forest cover change, two main land use activity trajectories were defined based on forest change, including afforestation and deforestation (Table 7), and the carbon budget caused by afforestation and deforestation was statistically analyzed. The results show that afforestation in Shaoguan led to net accumulation of vegetation biomass and soil carbon, resulting in an increase of 1.509 TgC in Shaoguan's forest carbon storage. Since the deforestation area in Shaoguan exceeded the afforestation area, the total carbon loss caused by deforestation was 2.802 TgC, far exceeding the increase in carbon storage brought about by afforestation. Figure 7 ).

[0114] In summary, from 2000 to 2022, the afforestation area in Shaoguan City was consistently smaller than the deforestation area. Even though afforestation can offset the carbon loss caused by deforestation to some extent, Shaoguan City's forests still showed a net carbon loss. This indicates that in order to achieve its dual carbon goals, Shaoguan still needs to carry out sustainable forest management and operation.

[0115] This invention proposes an improved method for estimating carbon budget of forest cover change by combining CASA and Bookkeeping models. This method not only obtains the spatial distribution of forest carbon density through the CASA model, but also estimates the response of forest carbon dynamics to forest cover change through the Bookkeeping model. This provides a scientific methodology for evaluating the impact of forest cover change on regional forest carbon dynamics. This method can also be applied to estimating carbon budget caused by land use / cover change.

[0116] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for estimating the carbon budget of forest cover change, characterized in that: By using the CASA model and the positive linear relationship between net primary productivity and forest biomass, medium- and high-resolution spatial distribution data of vegetation carbon density in the target area were obtained. Combining CASA and Bookkeeping models, based on the spatial distribution data of vegetation carbon density and soil carbon density, carbon flux estimation and distribution mapping of forest cover change with spatial details at the regional scale were realized to simulate more accurate regional carbon budget impact results of forest cover change. The specific implementation steps are as follows: S1: Collect and process Landsat and MODIS NDVI remote sensing data, land cover type data, monthly average temperature data, monthly total precipitation data, monthly total solar radiation data, and soil spatial distribution data; S2: Net primary productivity of the target region is simulated based on the CASA model and land cover type data; S3: Collect and process the average carbon density of each land type in the target area, and use the positive linear relationship between net primary productivity and forest biomass to calculate forest vegetation carbon density data; S4: Collect average soil carbon density data for the target area and calculate the change in soil organic carbon storage in the forest cover change area by multiplying litter productivity and litter turnover rate. S5: Estimate and analyze the carbon budget caused by forest cover change based on the improved Bookkeeping model.

2. The method for estimating carbon budget of forest cover change according to claim 1, characterized in that: In S1, the specific content of collecting and processing Landsat and MODIS NDVI remote sensing data, land cover type data, monthly average temperature data, monthly total precipitation data, monthly total solar radiation data, and soil spatial distribution data is as follows: The collected NDVI data were synthesized using a high-quality reconstruction algorithm that interpolates missing data from remote sensing time-series images to generate monthly-scale NDVI time-series data. We acquired monthly precipitation datasets and monthly average temperature datasets with 1 km resolution from the National Tibetan Plateau Scientific Data Center in China, and then interpolated them using the Kriging method to ensure that each dataset has a spatial resolution of 30 meters. We also generated monthly total solar radiation data from the long-sequence, high-density, and high-precision daily average solar radiation dataset based on station estimates, and interpolated the monthly total solar radiation data using the inverse distance weighting method to ensure that the data has a spatial resolution of 30 meters.

3. The method for estimating carbon budget of forest cover change according to claim 1, characterized in that: In S2, the net primary productivity of the target area, simulated based on the CASA model and land cover type data, is as follows: First, land cover type data are used to detect forest cover changes in the target area to obtain spatiotemporal distribution data of forest cover changes; Then, the model parameters were adjusted based on the maximum light energy utilization parameter to adapt them to the forests in the target area. Land cover type data, monthly NDVI time series data, monthly average temperature, monthly total precipitation and monthly total solar radiation data were imported into the CASA model to obtain the spatial distribution data of net primary productivity in the target area over the years.

4. The method for estimating carbon budget of forest cover change according to claim 1, characterized in that: In S3, the average carbon density of each land type in the target area is collected and processed, and the forest vegetation carbon density data is calculated using the positive linear relationship between net primary productivity and forest biomass. The specific details are as follows: The formula for calculating forest vegetation carbon density is as follows: Where i represents a pixel; C t N represents the average carbon density of forest vegetation. i C represents the net primary productivity of cell i; i Represents the forest vegetation carbon density of pixel i; n t This represents the total number of pixels occupied by forests in the target area; ∑ i∈t N i This represents the total net primary productivity of forest pixels.

5. The method for estimating carbon budget of forest cover change according to claim 1, characterized in that: In S4, the average carbon density data of soil types in the target area are collected, and the changes in soil organic carbon storage in the forest cover change area are calculated by multiplying litter productivity and litter turnover rate. By analyzing the litter decomposition process and employing a fixed first-order decomposition rate, the changes in soil organic carbon storage in forest cover change areas are calculated by multiplying litter productivity and litter turnover rate. The formulas for calculating changes in litter productivity and soil organic carbon storage are as follows: △SOCD=△Litter n ×TR L , Where n represents the nth year of each cycle, ΔLitter n NPP represents the litter productivity in year n. As and NPP Bs These refer to net primary productivity above and below ground, respectively; TR A and TR B These represent the turnover rates of aboveground and belowground biomass, respectively; ΔSOCD represents the change in soil organic carbon storage, and TR... L The turnover rate of litter biomass.

6. The method for estimating carbon budget of forest cover change according to claim 5, characterized in that: The reliability of the forest vegetation carbon density simulated by the CASA model was verified using field data from the target area in previous years. The specific details are as follows: To verify the accuracy of vegetation carbon density estimated from the CASA model, correlation statistical analysis was performed to compare the forest vegetation carbon density estimated by the CASA model in previous years with the measured forest vegetation carbon density data from the same year's small-plot survey. Sample points of forest plot data in the target area were selected by random sampling, and the aboveground biomass of the forest at the sample points was extracted and multiplied by the forest vegetation carbon content coefficient of 0.5 to convert it into forest vegetation carbon density data.

7. The method for estimating carbon budget of forest cover change according to claim 1, characterized in that: In S5, the specific content of estimating and analyzing the carbon budget caused by forest cover change based on the improved Bookkeeping model is as follows: The formula for calculating the carbon budget caused by forest cover change is as follows: Where M represents the carbon budget caused by forest cover change; m represents the total number of years studied; ΔSOC n ΔVC represents the change in soil carbon storage in year n. n This represents the change in vegetation carbon storage in year n. The formula for calculating changes in soil carbon storage caused by forest cover change is as follows: Wherein, ΔSOC n Let A represent the change in soil carbon storage in year n, where c and j represent land cover type c and another land cover type j, respectively, and the conversion between the two represents the change in forest cover type; A c,j,n It is the area where land cover type c changes to land cover type j in year n; ΔSOCD c,j,n This represents the change in soil organic carbon density when land cover type c changes to land cover type j in year n.

8. The method for estimating carbon budget of forest cover change according to claim 7, characterized in that: Changes in vegetation carbon storage caused by land cover change involve two main processes: the increase in vegetation organic carbon through restoration or artificial planting after the removal of native vegetation; and the release of removed vegetation biomass into the atmosphere as carbon dioxide at different oxidation rates after being used for construction, furniture making, or as carbon fuel. The calculation formula is as follows: △VC n =△VC RS -△VC RM , Wherein, ΔVC n It represents the change in vegetation carbon storage when a certain land cover type transforms into another land cover type in year n; ΔVC RS and ΔVC RM These represent the changes in vegetation carbon storage in year n due to biomass restoration and removal, respectively; A c,j,n It is the area where land cover type c changes to land cover type j in year n; VCD c Let α be the rate of change of vegetation carbon density for land cover type c. k It is the carbon storage of vegetation and in x k The ratio of total carbon reserves at different oxidation rates; x k This represents the oxidation rate of the removed vegetation under k different conditions; k represents the oxidation state of the removed vegetation.

Images