Method for calculating carbon stock of mixed forest ecosystem

Abstract

Provided is a method for calculating a carbon stock of a mixed forest ecosystem, which method belongs to the technical field of carbon sink estimation of forest ecosystems. The method comprises: acquiring basic geographic data, meteorological data, physiological and ecological parameters, historical operation data and verification data; on the basis of an existing Biome-BGC model, improving a phenological module, adding a thinning operation management module, and optimizing the physiological and ecological parameters, so as to propose an improved Biome-BGC model suitable for simulating a carbon stock of a managed mixed forest ecosystem; taking a mixed pine and oak forest as a research object, and using the improved model to simulate a carbon stock; verifying the improved model; using an extended Fourier amplitude sensitivity test method to analyze the sensitivity of the physiological and ecological parameters; and on the basis of selecting high-sensitivity parameters, using a path analysis method to analyze the influence of the high-sensitivity parameters on the carbon stock. The proposed improved model exhibits a better performance in the aspect of a carbon stock of a mixed forest.

Description

A method for calculating carbon storage in mixed forest ecosystems Technical Field

[0001] The present invention belongs to the technical field of forest ecosystem carbon sink estimation, and in particular relates to a method for calculating the carbon storage of a mixed forest ecosystem. Background Art

[0002] Forests are a major component of terrestrial ecosystems and play an important role in balancing global carbon regulation and mitigating the rise in atmospheric greenhouse gas concentrations. Forest ecosystems have a high carbon storage capacity, with their annual fixed carbon storage accounting for over 60% of the entire terrestrial ecosystem. Mixed forests, as complex ecosystems composed of multiple tree species, have greater potential for carbon storage, and their carbon storage is directly related to global carbon balance and climate change. Appropriate forest management (e.g., thinning and fertilization) can optimize forest structure and increase productivity, thereby enhancing the carbon storage potential of mixed forests. Accurately simulating the carbon storage of mixed forests under management is a strong guarantee for achieving sustainable forest development and carbon neutrality goals.

[0003] The Biome-BGC model is a widely used ecosystem process model that offers potential opportunities for studying the carbon cycle of terrestrial ecosystems and their responses to long-term and multi-scale disturbances. However, the Biome-BGC model can only simulate a single plant functional type within a single grid, such as evergreen coniferous forest (ENF) or deciduous broadleaf forest (DBF). For mixed forests, setting existing phenological models to either evergreen or deciduous forms presents several limitations. Previous simulation studies of mixed forests employed setting the phenological model to either evergreen or deciduous forms, followed by a weighted average based on their relative proportions (Li and Sun, 2018). While this approach retains the original model structure, its drawbacks include requiring the model to be run twice and, more importantly, leading to inaccurate descriptions of carbon cycling processes within mixed forests, resulting in significant errors in simulated carbon storage. Furthermore, human management practices can significantly impact the carbon cycle of forest ecosystems. Biome-BGC has been expanded to include management and disturbance practices, such as mowing and grazing in grasslands and bamboo shoot harvesting, selective felling, and fertilization in bamboo forests (Mao et al., 2016). Previous studies of thinning management typically defined the leaf area index (LAI) after thinning as a fixed value, thereby affecting the physiological and ecological processes of vegetation (Hidy et al., 2012). This approach only considers changes in leaves after thinning and fails to fully account for biomass losses in various plant organs. Biome-BGC has limitations in accurately quantifying carbon cycling processes under thinning management in mixed forests. Therefore, there is an urgent need to improve existing Biome-BGC models to effectively simulate carbon storage in managed mixed forest ecosystems.

[0004] Ecophysiological parameters of different forest types are key data necessary for the operation of the Biome-BGC model. The parameterization process is a crucial step when applying the Biome-BGC model. Currently, there are few studies on these ecophysiological parameters, and most rely on literature data to determine these parameters (Liu et al., 2022b). Due to insufficient prior knowledge of specific ecophysiological parameters of mixed forests, parameter uncertainty has an adverse impact on model accuracy. Many studies have used automatic calibration models such as simulated annealing (You et al., 2019), genetic algorithms (Miyauchi et al., 2019), and PEST (model-independent parameter estimation and uncertainty analysis) for global parameter optimization. Therefore, using observational data to apply optimization algorithms to optimize ecophysiological parameters to obtain a set of parameters suitable for simulating mixed forest carbon storage is an effective method to improve model accuracy.

[0005] Identifying the most influential parameters and considering the effects of interactions between parameters on carbon storage are crucial for accurate model calibration (Ren et al., 2022). This will provide strong support for improving model application and development. Sensitivity analysis is used to quantify the extent to which changes in model input parameters affect output. Some researchers have conducted global sensitivity analyses on carbon flux (Raj et al., 2014), carbon density of biomass pools (Miyauchi et al., 2019), and carbon storage (Robinson et al., 2009). Others have used machine learning methods to explore parameter sensitivity and uncertainty (Dagon et al., 2020). These studies have shown that sensitivity parameters vary across species and regions. However, relatively few analyses have examined how changes in sensitivity parameters affect simulation outputs in mixed forests. Therefore, identifying sensitive parameters that influence carbon storage and assessing their impact on carbon storage are key steps in simulating the carbon cycle in mixed forests.

[0006] Oak (Quercus, >400 species) and pine (Pinus, approximately 120 species) are the two major forest-forming genera in the temperate zone of the Northern Hemisphere. As a typical type of mixed forest, pine-oak mixed forests are widely distributed, primarily in subtropical to temperate and Mediterranean climate zones. They possess higher productivity, more stable ecosystems, and high carbon sequestration potential. Therefore, based on the existing Biome-BGC model, by improving the phenological module, adding a thinning and management module, and optimizing physiological and ecological parameters, an improved Biome-BGC model suitable for simulating carbon storage in managed mixed forest ecosystems was proposed. Using a pine-oak mixed forest as the research object, the improved model was used to simulate carbon storage and validate the model. Furthermore, identifying ecophysiological parameters with high sensitivity to carbon storage and evaluating their impact on carbon storage are important for accurately simulating the carbon cycle in mixed forests.

[0007] Summary of the Invention

[0008] In view of the above problems, the technical problem to be solved by the present invention is to provide a method for calculating carbon storage in mixed forest ecosystems, improve the Biome-BGC model and optimize physiological and ecological parameters to enhance the applicability of the model in managed mixed forest ecosystems, explore the sensitive physiological and ecological parameters that affect the carbon storage of mixed forests, and evaluate the impact of highly sensitive physiological and ecological parameters on the carbon storage of mixed forests.

[0009] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0010] A method for calculating carbon storage in a mixed forest ecosystem comprises the following steps:

[0011] S1: Acquire basic geographic data, meteorological data, physiological and ecological parameters, management history data and verification data;

[0012] S2: Based on the existing Biome-BGC model, the phenological module is improved, a thinning management module is added, and physiological and ecological parameters are optimized to propose an improved Biome-BGC model for carbon storage in mixed forest ecosystems suitable for simulated management. Specifically, the model includes:

[0013] S201: Improve the phenology module based on the evergreen phenology model to develop a phenology model suitable for simulating mixed forests. Specifically:

[0014] S2011: Define the parameter as the ratio of the transition growth period to the deciduous vegetation growing season to calculate the start and end time of the deciduous vegetation transition period;

[0015] S2012: The parameter was defined as the ratio of evergreen vegetation to deciduous vegetation to calculate the daily transfer amount in different phenological periods of mixed forests;

[0016] S2013: Define the parameter as the proportion of litter fall during the growing season of deciduous vegetation, describe the start and end time of the litter fall process of deciduous vegetation, and calculate the daily litter fall in different phenological periods of mixed forests based on the proportion parameter of evergreen vegetation and deciduous vegetation.

[0017] S202: Add a thinning management module to simulate the impact of thinning on mixed forest carbon storage;

[0018] S203: Optimize and analyze physiological and ecological parameters using a pollination algorithm to establish a set of parameters suitable for simulating carbon storage in mixed forests;

[0019] S3: Taking mixed forest as the research object, the improved Biome-BGC model was used to simulate carbon storage;

[0020] S4: Validation of the improved Biome-BGC model;

[0021] S5: Analysis of the sensitivity of physiological and ecological parameters using the extended Fourier amplitude sensitivity test method;

[0022] S6: Based on the selection of highly sensitive parameters, the path analysis method is used to analyze the positive and negative effects of highly sensitive parameters on carbon storage.

[0023] Preferably, the basic geographic data include: digital elevation model, slope, aspect, soil sand, clay and silt content, shortwave albedo, nitrogen deposition and nitrogen fixation.

[0024] The meteorological data include: daily maximum temperature, daily minimum temperature, daily average temperature, daily precipitation, daily saturated vapor pressure difference, daily shortwave radiation flux density and day length.

[0025] The physiological and ecological parameters are as follows: on the basis of the physiological and ecological parameters defined in the original model, three parameters are added to calculate the amount of transfer and litter, including the ratio of evergreen vegetation to deciduous vegetation, the ratio of the transfer growth period to the deciduous vegetation growth season, and the ratio of the littering process to the deciduous vegetation growth season.

[0026] The management history data: 10 parameters are defined in the thinning management module, including the thinning date, the harvesting rate of each plant organ, and the transportation rate.

[0027] The verification data: verification is performed using measured data.

[0028] Preferably, in S2011, the parameter is defined as the ratio of the transition growth period to the deciduous vegetation growing season to calculate the start and end time of the deciduous vegetation transition period, specifically:

[0029] For deciduous vegetation, the transition period is described based on the start and end of the growing season. Leaf unfurling occurs when the sum of the daily mean soil temperatures exceeds a defined threshold. The actual date of leaf unfurling is 15 days earlier than this calculated date, at which point the growing season begins: onset day =m(ST soil ≥T crit ) Act onset_day =pmset day -15

[0030] Among them, Tsoil_avg i is the average soil temperature on the i-th day of the year, ST soil is the sum of the daily average soil temperatures when the average soil temperature is greater than 0, T avg is the average daily temperature during the operation days, Tcrit is the defined critical value, onset day is the calculated leaf unfolding day, Act onset_day is the actual leaf unfolding date; m is the time when ST soil Greater than or equal to T crit That day when

[0031] If it is after July 1st and the day length is less than 10 hours and 55 minutes; and the soil temperature is lower than the average soil temperature in autumn or the soil temperature is lower than 2°C, all leaves will fall. The actual date of leaf fall is 15 days later than the calculated date, at which time the growing season ends;

[0032] All leaves fall when one of the following conditions is met: offset day =j Actoffset_day=offset day +15

[0033] Among them, Day j is the length of the day j in a year, Tsoil_avg j is the average soil temperature on the jth day of the year, Tsoil avg_aut is the average soil temperature between September and October, offset day Actoffset_day is the calculated day when all leaves fall off, and Actoffset_day is the actual day when all leaves fall off. The growing season is calculated based on the actual leaf expansion day and the actual leaf fall day; ngrowthdays = Actoffset_day - Act onset_day t1=Act onset_day t2=Actoffset_day+ngrowthdays×T t_d

[0034] Where ngrowthdays is the number of days in the growing season, t1 and t2 are the start and end days of the transition period in deciduous vegetation, respectively, and T t_d It is the proportion of the transition growth period to the growing season of deciduous vegetation.

[0035] As a preferred method, in S2012, the parameter defined as the ratio of evergreen vegetation to deciduous vegetation is used to calculate the daily transfer amount in different phenological periods of the mixed forest, specifically:

[0036] Among them, S daily_transfer is the daily transfer volume of mixed forest, C transferis the transfer amount of each plant organ in the mixed forest, ndays_E is the remaining transfer days of evergreen vegetation, ndays_D is the remaining transfer days of deciduous vegetation, E:D is the ratio of evergreen vegetation to deciduous vegetation, and nday represents the day of the year.

[0037] As a preferred embodiment, S2013: define the parameter as the ratio of the littering process to the growing season of deciduous vegetation, describing the start and end of the littering process of deciduous vegetation, specifically: t3 = Actoffset_day-(ngrowthdays×LFG d )+1 t4=Actoffset_day

[0038] Among them, t3 and t4 are the start and end dates of the littering process in deciduous vegetation, respectively. d is the proportion of littering process in the growing season of deciduous vegetation;

[0039] The daily litter fall in mixed forests at different phenological stages was calculated based on the ratio parameters of evergreen vegetation and deciduous vegetation. Specifically:

[0040] Among them, S daily_litterfall is the daily litterfall amount, Clitterfall_increment_E is the daily litterfall amount of evergreen plants, which is a constant rate throughout the year, Clitterfall_increment_D is the daily litterfall amount of deciduous vegetation, C annmax is the annual maximum daily carbon, F_turnover is the annual turnover rate, C leaf_froot Is the carbon content in leaves or rootlets, litdays D is the number of days left for deciduous vegetation to fall, Drate is the linear growth rate, t represents the number of days required to remove all fine roots and leaves and the number of iterations in the equation, is the daily litterfall of deciduous vegetation on day t, is the daily litterfall of deciduous vegetation on day t+1.

[0041] Preferably, in S202, a thinning management module is added to simulate the impact of thinning on the carbon storage of mixed forests, specifically:

[0042] Define the parameters of thinning date, harvest rate and transport rate of each plant organ. The harvest rate is the proportion of biomass removed during thinning, and the transport rate is the proportion removed from the site after thinning and no longer participates in the carbon cycle after transportation.

[0043] The carbon loss and carbon entering the next litter pool of each plant organ are calculated based on the thinning rate and transportation rate, specifically: C all_to_THN =C all ×THN

[0044] C all_to_THN_litr =C all_to_THN ×(100-TRAN) / 100×litter

[0045] Where THN is the harvesting rate, C all is the carbon content of each organ, C all_to_THN The carbon flux generated by thinning in each organ is considered as the carbon loss of each organ. TRAN is the transport rate of each organ, litter is the proportion of each organ entering the four litter pools, and C all_to_THN_litr Describes the carbon that is converted to the corresponding litter components after thinning and remains in situ.

[0046] Preferably, in S203, the flower pollination algorithm is used to optimize and analyze the physiological and ecological parameters to establish a set of parameters suitable for simulating carbon storage in mixed forests, specifically:

[0047] The maximum number of iterations, the number of individuals in the population, and the transition probability are set. The objective function is the minimum residual sum between the simulated carbon storage value and the measured value. The decision variable is the physiological and ecological parameter to be optimized. The parameter value is repeatedly adjusted within the feasible range until the objective function reaches the ideal minimum value.

[0048] As a preferred method, in S3, the mixed forest was used as the research object and the improved Biome-BGC model was used to simulate carbon storage:

[0049] In the first step, the spin-up process is used to bring the model into a stable state, and the model outputs a restart file; in the second step, the model uses the restart file as input and runs forward from this point on, starting the thinning management module in the corresponding areas where thinning management is adopted to simulate the impact of thinning on carbon storage.

[0050] Preferably, in S4, the improved Biome-BGC model is verified by:

[0051] In the first step, the original model's phenological module was set to simulate evergreen phenology and deciduous phenology separately. Then, a weighted average was taken based on the proportion of mixed forest. This simulation result was compared with the simulation result of the improved model adjusted by the phenological module.

[0052] The second step was to compare the simulation results of the improved model adjusted by the phenological module with the simulation results of the improved model that adjusted the phenological module and added the thinning module. All simulation processes were simulated using optimized eco-physiological parameters.

[0053] By calculating R 2 , MAE and RMSE evaluate the simulation results of the improved Biome-BGC model, and the calculation formula is as follows:

[0054] Among them, n represents the number of measured data, R 2 reflects the degree of fitting, MAE and RMSE reflect the degree of difference, R 2 The closer it is to 1, the lower the MAE and RMSE values ​​are, which means the simulation accuracy is higher.

[0055] Preferably, in S5, the method for analyzing the sensitivity of the physiological and ecological parameters using the extended Fourier amplitude sensitivity test method is as follows: Y = f(X) = f(X1, X2, X3, ..., X n ) S i =V i / V Y S iT =S i +S ij +S ijk +…+S 12…n

[0056] Among them, Y is the output result of the improved Biome-BGC model, X i are the physiological and ecological parameters of a given distribution range, V Y is the total variance of the output, V i is the variance of a single parameter, V ij -V 12…n is the variance of the interaction between parameters, S i It is the first-order sensitivity index, which indicates the direct contribution of the parameter to the total variance of the output result. iT is the global sensitivity index, which is the sum of the first-order sensitivity of the parameter and the sensitivity indices of each order of the interaction between the parameter and other parameters.

[0057] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0058] (1) This paper modifies the phenology module by introducing the growing season of deciduous vegetation into the evergreen phenology module and calculating the migration and littering processes of mixed forests at different periods. The optimized phenology module helps to comprehensively capture the dynamic changes of mixed forest ecosystems, thereby improving the understanding of internal ecosystem processes and enhancing the accuracy of the Biome-BGC model's simulation results for ecological processes such as the carbon cycle in mixed forests.

[0059] (2) Based on the improved phenological module, the present invention incorporates a thinning management module into the model, which helps to simulate the impact of thinning on the carbon cycle of mixed forests and the changes in carbon reserves under different thinning management scenarios in the future, and helps to formulate sustainable forest management plans and balance economic and ecological benefits.

[0060] (3) This paper optimizes the ecophysiological parameters of mixed forests to establish a set of parameters suitable for simulating carbon storage in mixed forests. This overcomes the uncertainty in the values ​​of these parameters, thereby enhancing the versatility and practicality of the model. The optimized ecophysiological parameters contribute to a deeper understanding of the function and structure of mixed forest ecosystems, thereby improving the accuracy of the model's simulation of ecological processes in mixed forests.

[0061] (4) The present invention can understand which physiological and ecological parameters are most sensitive to the impact of simulation results, allowing for more precise parameter calibration, thereby improving model accuracy. This helps to develop more effective ecosystem management strategies to maximize the maintenance or improvement of carbon storage.

[0062] (5) The present invention can evaluate the positive and negative effects of highly sensitive physiological and ecological parameters on the carbon storage of mixed forests, provide an in-depth understanding of the contribution of each parameter to carbon storage, provide a theoretical basis for subsequent model parameter optimization and parameter calibration, and further deepen the understanding of the internal processes and dynamics of mixed forest ecosystems. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 is a schematic diagram of the experimental area; (a) detailed characteristics of the Qinling Mountains, and (b) thinning sites and carbon storage verification sites in the Qinling Mountains (Note: T + year indicates thinning management carried out in the specified year, for example, T2006 indicates thinning carried out in 2006);

[0064] FIG2 is a flow chart of a method for calculating carbon storage in a mixed forest ecosystem;

[0065] FIG3 is a flow chart of a carbon storage calculation method S2 for a mixed forest ecosystem;

[0066] FIG4 is a flow chart of a method S201 for calculating carbon storage in a mixed forest ecosystem;

[0067] Figure 5 is a conceptual diagram of the improved Biome-BGC model for simulating carbon cycle in mixed forests;

[0068] Figure 6 is a validation diagram of the improved model with modified phenology module (1:1 linear relationship, simulations using the original model and the improved model are represented in black, green and red, respectively.);

[0069] Figure 7 is a validation diagram of the improved model that modifies the phenology module and introduces the thinning module (the 1:1 linear relationship, the simulation using the original model and the improved model are represented in black, green and red, respectively.);

[0070] Figure 8 is a schematic diagram of the first-order sensitivity index and total sensitivity index of carbon storage;

[0071] Figure 9 is a diagram of the path analysis model between highly sensitive parameters and carbon storage. DETAILED DESCRIPTION

[0072] The present invention will be further illustrated below with reference to specific examples. The examples are implemented based on the technical solutions of the present invention. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0073] As shown in Figure 1, the present invention uses the Qinling Mountains as the experimental area and the pine-oak mixed forest as the research object. The Qinling Mountains cover an area of ​​approximately 63,700 km 2 The Qinling Mountains are located primarily in southern Shaanxi Province, China (32°28'N-34°40'N, 105°28'E-113°3'E). As a key geographic dividing line between north and south my country, it serves as a transitional zone between northern and southern vegetation. The northern region experiences a warm temperate semi-humid climate with an annual precipitation of 520 mm and an annual temperature of 10°C, while the southern region experiences a subtropical humid climate with an annual precipitation of 820 mm and an annual temperature of 14°C. Below 4,000 m above sea level, the terrain is undulating, gradually increasing from east to west. As the most extensive forest ecosystem in central China, the Qinling Mountains play a key role in forest and ecological conservation. Pinus tabulaeformis and Quercus acuteseri are the dominant species in the widespread pine-oak mixed forests of the Qinling Mountains, with a wide distribution range and important role in the ecosystem. From 2000 to 2019, low-intensity thinning management, a key conservation management practice, was implemented to maintain ecological stability and restore the zonal climax community.

[0074] As shown in Figures 2-3, the present invention proposes a method for calculating carbon storage in a mixed forest ecosystem, comprising the following steps:

[0075] S1: Acquire basic geographic data, meteorological data, physiological and ecological parameters, management history data and verification data;

[0076] (1) Basic geographic data: digital elevation model (DEM), slope, aspect, soil sand, clay, and silt content, shortwave albedo, nitrogen deposition, and nitrogen fixation. The digital elevation model (DEM) was provided by the Resource and Environmental Science Data Center of the Chinese Academy of Sciences (https: / / www.resdc.cn / ). Slope and aspect data were derived from the DEM data using ArcGIS, with a resolution of 1 km. The Chinese soil dataset (v1.1) (2009) of the Harmonized World Soil Database (HWSD) (http: / / westdc.westgis.ac.cn / zh-hans) was used. The data source for China was the 1:1 million soil data provided by the Second National Land Survey in Nanjing (FAO, 2019), with a resolution of 1 km. This dataset includes data on soil sand, clay, and silt content and soil organic carbon content. Shortwave albedo was obtained from flux tower measurements. Nitrogen deposition and nitrogen fixation used the default model data.

[0077] (2) Meteorological data: Daily meteorological variables mainly consist of seven indicators: daily maximum temperature, daily minimum temperature, daily mean temperature, daily precipitation, daily saturated vapor pressure difference, daily shortwave radiation flux density, and day length. The daily maximum temperature, daily minimum temperature, and daily precipitation from 1980 to 2019 were provided by the China Meteorological Data Platform (http: / / data.cma.cn / ) and interpolated using the ANUSPLIN software. Other indicators were inverted using the MT-CLIM model (Mountain Microclimate Simulation Model) based on the above three variables. Atmospheric carbon dioxide concentration was obtained from the observation dataset of the National Oceanic and Atmospheric Administration (NOAA, http: / / www.noaa.gov / web.html). All spatial data resolutions were resampled to 1 km, and the geographic coordinate system was unified to WGS 1984 for consistent analysis.

[0078] (3) Physiological and ecological parameters: The model requires 44 physiological and ecological parameters in the original model. Three new parameters are defined in the improved model to calculate the amount of transfer and litter; namely, the ratio of evergreen vegetation to deciduous vegetation, the ratio of the transfer growth period to the deciduous vegetation growing season, and the ratio of the litter process to the deciduous vegetation growing season. Some of these parameters do not change with plant functional type, and the other 40 parameters need to be adjusted according to the pine-oak mixed forest. Most of the parameters are set based on previous studies, but due to lack of data, 17 parameters need to be optimized through optimization algorithms to obtain a set of physiological and ecological parameters suitable for pine-oak mixed forests. Since the three parameters (the ratio of the transfer growth period to the evergreen vegetation growing season, the ratio of the litter process to the evergreen growth season, and the annual fire mortality rate) are all fixed values, no sensitivity analysis is required. Table 1 below shows the physiological and ecological parameters of the pine-oak mixed forest ecosystem.

[0079] Table 1 Physiological and ecological parameters of pine-oak mixed forest ecosystem

[0080] (4) Historical management data: Ten new parameters are defined in the thinning management module, mainly including thinning date, harvesting rate and transportation rate of various plant organs, etc. The relevant data are from previous research and forest tending operation design data of the Forestry Bureau, as shown in Table 2.

[0081] Table 2 Historical operating data

[0082] (5) Validation data: For vegetation carbon storage, 485 field data from the 6th (2003), 7th (2008), and 8th (2013) forest resource inventory data in the Qinling Mountains were used to optimize the physiological and ecological parameters. The model was validated using data from the 9th (2018) forest resource inventory. For soil carbon storage, field data were obtained from the Soil Center of the National Earth System Science Data Center (http: / / soil.geodata.cn) (Liu Feng, 2020). For litter carbon storage, some field data are shown in Table 3 below.

[0083] Table 3 Litter carbon storage verification data

[0084] S2: Based on the Biome-BGC model version 4.2, the phenological module was improved, a thinning management module was added, and the physiological and ecological parameters were optimized to propose an improved Biome-BGC model for carbon storage in mixed forest ecosystems suitable for simulated management.

[0085] To improve the applicability of Biome-BGC in mixed forests, the present invention assumes that the phenological model is an evergreen model before simulation, and then introduces the growing season of deciduous forest into the evergreen phenological module to develop a phenological module suitable for mixed forests.

[0086] Mixed forest ecosystems are disturbed by thinning management. Therefore, the present invention introduces a thinning management module into the model to simulate the impact of thinning on the carbon storage of mixed forests.

[0087] To address the uncertainty of the specific values ​​of mixed forest ecophysiological parameters, the present invention combines measured data with an optimization algorithm to optimize these parameters. Through optimization, the present invention establishes a set of optimized ecophysiological parameters suitable for simulating carbon storage in mixed forests.

[0088] S201: Improve the phenology module based on the evergreen phenology model, specifically:

[0089] S2011: Define the parameter as the ratio of the transition growth period to the deciduous vegetation growing season to calculate the start and end time of the deciduous vegetation transition period;

[0090] During the growing season, the existing Biome-BGC model's daily phenology transfers carbon to new tissues. The model's evergreen vegetation transition period is year-round. This paper incorporates the transition period of deciduous vegetation into the evergreen model and updates the daily transfer amounts for different phenological periods based on the ratio of evergreen to deciduous vegetation in mixed forests.

[0091] For deciduous vegetation, the transition period is described based on the start and end of the growing season. soil >T crit ) when the leaves begin to unfold. The actual date of leaf unfolding is 15 days earlier than the calculated date, which is the beginning of the growing season: onset day =m(ST soil ≥T crit ) Act onset_day =onset day -15

[0092] Among them, Tsoil_avg i is the average soil temperature on the i-th day of the year, ST soil is the sum of the daily average soil temperatures when the average soil temperature is greater than 0, T avg is the average daily temperature during the operation days, T crit is the defined critical value, onset day is the calculated leaf unfolding day, Act onset_day It is the actual leaf unfolding day. m represents Dangdang ST soil Greater than or equal to T crit That day.

[0093] If it is after July 1st, the day length is less than 10 hours and 55 minutes (39,300 seconds), and the soil temperature is below the average soil temperature in autumn (September and October in the Northern Hemisphere) or below 2°C, all leaves will fall. The actual date of leaf fall is 15 days later than this calculated date, which is the end of the growing season.

[0094] All leaves fall when one of the following conditions is met: offset day =j Actoffset_day=offset day +15

[0095] Among them, Daylen j Indicates the length of the jth day in a year, Tsoil_avg j is the average soil temperature on the jth day of the year, Tsoil avg_aut is the average soil temperature between September and October, offset day Actoffset_day is the calculated day when all leaves fall, and Actoffset_day is the actual day when all leaves fall. The growing season is calculated based on the actual leaf expansion day and the actual leaf fall day.

[0096] This embodiment defines a new parameter, namely, the ratio of the transition growth period to the deciduous vegetation growth season (T t_d , see Table 1) to calculate the end time of the deciduous vegetation transition period: ngrowthdays = Actoffset_day - Act onset_day t1=Act onsey_day t2=Actoffset_day+ngrowthdays×T t_d

[0097] where ngrowthdays represents the number of days in the growing season, and t1 and t2 are the start and end days of the transition period in deciduous vegetation, respectively.

[0098] S2012: The parameter was defined as the ratio of evergreen vegetation to deciduous vegetation to calculate the daily transfer amount in different phenological periods of mixed forests;

[0099] The present invention defines the ratio of evergreen vegetation to deciduous vegetation (E:D, see Table 1) to calculate the daily transfer volume in different phenological periods of mixed forests. The calculation formula is as follows:

[0100] Among them, S daily_transfer is the daily transfer volume of mixed forest, C transferis the transfer amount of various plant organs (including leaves, fine roots, living stems, dead stems, living thick roots and dead thick roots) in the mixed forest, ndays_E is the remaining transfer days of evergreen vegetation, ndays_D is the remaining transfer days of deciduous vegetation, and nday represents the day of the year.

[0101] S2013: The parameter was defined as the proportion of littering in the growing season of deciduous vegetation, describing the start and end time of the littering process of deciduous vegetation, and obtaining the daily littering amount in different phenological periods of mixed forests;

[0102] Litterfall transfers carbon from fine roots and leaves to the four litter pools in the proportions specified in Table 1. Evergreen vegetation produces litter daily throughout the year, while the litterfall process of deciduous vegetation exhibits significant seasonal variation. Therefore, the model defines the litterfall cycle for deciduous vegetation and then calculates daily litterfall based on the ratio of evergreen to deciduous vegetation in the mixed forest.

[0103] The present invention defines a new parameter, namely the proportion of littering process in the growing season of deciduous vegetation (LFG d , see Table 1) to describe the start and end of the fall process. t3 = Actoffset_day-(ngrowthdays×LFG d )+1 t4=Actoffset_day

[0104] Among them, t3 and t4 are the starting and ending dates of the littering process in deciduous vegetation, respectively.

[0105] The daily litterfall amount in different phenological periods of mixed forests is calculated as follows:

[0106] Among them, S daily_litterfall is the daily litterfall, Clitterfall_increment_E is the daily litterfall of evergreen plants (including leaves and fine roots), which is a constant rate throughout the year, Clitterfall_increment_D is the daily litterfall of deciduous vegetation, which starts from 0 and increases at a linear growth rate (Drate) so that all fine roots and leaves have fallen before t4, and C annmax is the annual maximum daily carbon, F_turnover is the annual turnover rate, C leaf_froot Is the carbon content in leaves or rootlets, litdays D is the number of days left for deciduous vegetation to fall off; t represents the number of days required to remove all fine roots and leaves and the number of iterations in the equation, is the daily litterfall of deciduous vegetation on day t, is the daily litterfall of deciduous vegetation on day t+1.

[0107] S202: Add a thinning management module to simulate the impact of thinning on mixed forest carbon storage;

[0108] The present invention defines the following parameters: (1) thinning date (e.g. April 20, 2001), (2) harvesting rate of each plant organ, which refers to the proportion of biomass removed during thinning, calculated as a percentage. It mainly includes the harvesting rate of leaves, living stems, dead stems, living thick roots, dead thick roots and fine roots. (3) transportation rate, which refers to the proportion of biomass removed from the site after thinning and no longer participates in the carbon cycle after transportation is completed, calculated as a percentage. The parameters include the transportation rate of leaves, living stems and dead stems. In addition, the present invention defines that living thick roots, dead thick roots and fine roots are not transported, remain on site and become part of dead wood or litter pool according to the root type.

[0109] The carbon loss and carbon entering the next litter pool of each plant organ are calculated based on the thinning rate and transportation rate, specifically: C all_to_THN =C all ×THN C all_to_THN_litr =C all_to_THN ×(100-TRAN) / 100×litter

[0110] Where THN is the harvesting rate, C all is the carbon content of each organ, C all_to_THN describes the carbon flux of each organ due to thinning, which is considered as the carbon loss of each organ. TRAN is the transport rate of each organ, and litter is the proportion of each organ entering the four litter pools, which are specified in Table 1. all_to_THN_litr Describes the carbon that remains on site due to thinning and is converted into the corresponding litter components. After the above functions are completed, the actual vegetation carbon pool and litter carbon pool will be modified and updated.

[0111] S203: Optimize and analyze physiological and ecological parameters using a pollination algorithm to establish a set of parameters suitable for simulating carbon storage in mixed forests;

[0112] A mixed pine and oak forest was used as the experimental object. The Flower Pollination Algorithm (FPA) was used to optimize and analyze the 17 physiological and ecological parameters listed in Table 1. This algorithm is a popular intelligent optimization algorithm with the advantages of high speed and resistance to local extremes. The present invention sets the maximum number of iterations N = 50, the number of individuals in the population n = 20, and the transition probability p = 0.8. The objective function is the minimum residual sum between the simulated carbon storage value and the measured value, and the decision variables are the 17 physiological and ecological parameters. The values ​​of the parameters are repeatedly adjusted within the feasible range until the objective function reaches the ideal minimum value, thus completing the optimization of the physiological and ecological parameters.

[0113] S3: Using pine-oak mixed forest as the research object, the improved Biome-BGC model was used to simulate carbon storage;

[0114] The improved Biome-BGC model operates in two phases. The first is a spin-up process to bring the model to a steady state, and the model outputs a restart file. This typically requires steady-state initial conditions to ensure a balance between input and output fluxes and that the system is in equilibrium with the environment. The second phase involves the model using the restart file as input and running forward from this point. The thinning management module is activated in the corresponding areas where thinning measures have been implemented to simulate the impact of thinning on carbon storage.

[0115] The Biome-BGC model is a one-dimensional model that simulates at a point level. The present invention utilizes code compilation to perform grid-by-grid simulation, thereby achieving regional simulation. The carbon pools output by the present invention include vegetation carbon storage, litter carbon storage, soil carbon storage, and total carbon storage.

[0116] S4: Validation of the improved Biome-BGC model;

[0117] As shown in Figures 6 and 7, in the first step, the original model's phenological module was set to simulate evergreen and deciduous phenology separately. A weighted average was then taken based on the proportion of mixed forest. This simulation result was compared with that of an improved model adjusted with the phenological module. In the second step, the simulation results of the improved model adjusted with the phenological module were compared with those of an improved model that also adjusted the phenological module and added a thinning module. All simulations were performed using optimized ecophysiological parameters.

[0118] The present invention calculates R 2 , MAE (mean absolute error) and RMSE (root mean square error) are used to evaluate the simulation results. The calculation formulas are as follows:

[0119] Among them, n represents the number of measured data, R 2 It can reflect the degree of fitting, while MAE and RMSE can reflect the degree of difference. 2 The closer it is to 1, the lower the MAE and RMSE values ​​are, which means the simulation accuracy is higher.

[0120] S5: Analysis of the sensitivity of physiological and ecological parameters using the extended Fourier amplitude sensitivity test method;

[0121] The present invention uses the Extended Fourier Amplitude Sensitivity Test (EFAST) method to analyze the sensitivity of physiological and ecological parameters. EFAST is a variance-based global sensitivity analysis algorithm that combines the computational efficiency of the FAST (Features from accelerated segment test) method with the advantages of the overall sensitivity of the Sobol's method (Sobol, 1993). It can quantify the sensitivity of each parameter and the interaction between parameters to the results. It is widely used in the sensitivity analysis of nonlinear models, such as hydrological models, crop growth models, Biome-BGC, etc. By estimating the sensitivity of each input parameter (X i ) to the variance contribution of the simulation results (Y), which can be calculated by the following equation: Y = f(X) = f(X1, X2, X3, ..., X n )

[0122] Among them, Y is the output of the improved Biome-BGC model (vegetation carbon storage, soil carbon storage, litter carbon storage and total carbon storage), X i Represents various physiological and ecological parameters of a given distribution range.

[0123] The total variance of the model output can be expressed as:

[0124] Among them, V Y is the total variance of the output, V i is the variance of a single parameter, V ij -V 12…n is the variance of the interaction between parameters.

[0125] Sensitivity is measured by the contribution of a given input factor to the variance of the output result. i ) and the global sensitivity index (S iT ) to quantify the contribution of the input parameters to the output results. The first-order sensitivity index represents the direct contribution of the parameter to the total variance of the output results and is calculated by the following equation: S i =V i / V Y

[0126] The global sensitivity index is the sum of the first-order sensitivity of the parameter and the sensitivity index of each order of the interaction between the parameter and other parameters. The calculation formula is as follows: S iT =S i +S ij +S ijk +…+S 12…n

[0127] In the present invention, 37 parameters were first selected for sensitivity analysis based on actual conditions (see Table 1). According to the distribution range of each input parameter, the Monte Carlo method was used to randomly sample each parameter, and the sampling frequency of each parameter was set to 130 times. Therefore, the sampling times of the physiological and ecological parameters were 4810 (130×37) times. According to the generated multiple sets of input parameters, Biome-BGC was run in batches to simulate the carbon storage of pine-oak mixed forests from 1980 to 2019, and the average carbon storage for 40 years was calculated as the final model input. The sensitivity and uncertainty analysis software SimLab2.2 was used to analyze the sensitivity of the physiological and ecological parameters in Biome-BGC. Finally, the sensitivity index was divided into three levels, greater than 0.2 was a highly sensitive parameter, 0.1-0.2 was a medium sensitive parameter, and less than 0.1 was an insensitive parameter.

[0128] S6: Based on the selection of highly sensitive parameters, the path analysis method was used for statistical analysis to obtain the results.

[0129] Based on the selection of highly sensitive parameters, path analysis was used to explore the impact of parameters and their interactions on the output. Path analysis is a multivariate statistical analysis method that can reflect the causal relationship between variables in a model. The path coefficient can reflect the strength of the causal relationship between variables. In this paper, path analysis was performed using the lavaan package in the R language.

[0130] The optimization results of 17 physiological and ecological parameters of pine-oak mixed forest are shown in Table 4.

[0131] Table 4 Optimal values ​​of 17 physiological and ecological parameters in pine-oak mixed forests

[0132] As shown in Figure 6, the simulation accuracy of the improved model with the modified phenology module was significantly improved in all three carbon components (vegetation carbon, litter carbon, and soil carbon) compared with the simulation using the original model. 2 The values ​​increased by 0.32, 0.21, and 0.46, respectively, with vegetation carbon and soil carbon increasing by more than twofold. Similarly, the reductions in MAE and RMSE were significant (20-76%). Furthermore, the new model improved the slope of the 1:1 linear relationship between the measured and simulated data for the three carbon pools. The simulation results of the improved model with modified phenology modules are reasonable and can be applied to simulate carbon storage in mixed pine-oak forests.

[0133] As shown in Figure 7, due to the lack of verification data for litter carbon storage, the present invention uses the verification of plant carbon storage and soil carbon storage to evaluate the simulation accuracy of the improved model with modified phenology module and the improved model with modified phenology module and thinning module. In the improved model with modified phenology module, the R between the simulated value and the measured value of vegetation carbon storage is 0. 2 , MAE and RMSE are 0.43, 1.09 and 1.43 kg C / m respectively 2 , soil carbon storage was 0.55, 0.10 and 0.13 kg C / m 2 In the improved model that modified the phenological module and introduced the thinning module, the R 2 , MAE and RMSE are 0.62, 0.93 and 1.16 kg C / m 2 , soil carbon storage was 0.65, 0.09 and 0.11 kg C / m 2 The improved model that modified both the phenological module and the thinning module performed better in simulating vegetation carbon storage and soil carbon storage than the improved model that only modified the phenological module.

[0134] Figure 8 shows the sensitivity analysis of carbon storage. High sensitivity parameters are above the red dashed line, and medium sensitivity parameters are above the blue dashed line and below the red dashed line. i and S iT The highest, through direct or interactive control of carbon storage. Vegetation carbon storage in S i There is no highly sensitive parameter in S, and it is most affected by WPM. iT WPM, LFRT, WPM, SC:LC, CRC:SC, C:N litter , SLA and G cut It plays a significant role in controlling the four carbon pools. iT Than S i There were more sensitivity parameters, which indicated that many parameters affected carbon storage only through interactions, especially in plant carbon storage.

[0135] As shown in Figure 9, the effects of physiological and ecological parameters on carbon storage, for litter carbon storage, soil carbon storage and total carbon storage, G cut and SLA have a direct positive effect, among which G cut The impact is the greatest. WPM has a large indirect positive effect through SLA. The direct impact of LFRT, WPM and SC:LC are all negative, with LFRT showing the highest performance. cut and SLA both have a greater indirect negative effect through each other. C:N litterLFRT has a significant indirect negative effect on soil carbon storage and total carbon storage. litter , SLA and T t_d showed a direct positive effect, while LFRT and WPM showed a direct negative effect. SLA had a larger indirect negative effect through WPM.

[0136] In Figure 9 , all path coefficients have been standardized. Solid lines represent direct effects, and dashed lines represent indirect effects. Green lines represent positive effects, and red lines represent negative effects. The width of the arrows is proportional to the size of the standardized path coefficients (P<0.01).

Claims

1. A method for calculating the carbon storage of a mixed forest ecosystem, characterized in that, It includes the following steps: S1: Obtain basic geographical data, meteorological data, physiological and ecological parameters, operation history data, and verification data; S2: Based on the existing Biome - BGC model, improve the phenology module, add a thinning operation management module, and optimize the physiological and ecological parameters to propose an improved Biome - BGC model applicable to simulating the carbon storage of managed mixed - forest ecosystems. Specifically, it includes: S201: Improve the phenology module based on the evergreen phenology model to develop a phenology model suitable for simulating mixed forests. Specifically: S2011: Define a parameter as the proportion of the transition growth period in the growth season of deciduous vegetation to calculate the start and end times of the transition period of deciduous vegetation; S2012: Define a parameter as the proportion of evergreen vegetation and deciduous vegetation to calculate the daily transfer amount during different phenological periods of the mixed forest; S2013: Define a parameter as the proportion of the litter - fall process in the growth season of deciduous vegetation to describe the start and end times of the litter - fall process of deciduous vegetation, and calculate the daily litter - fall amount during different phenological periods of the mixed forest according to the proportion parameter of evergreen vegetation and deciduous vegetation; S202: Add a thinning operation management module to simulate the impact of thinning on the carbon storage of the mixed forest; S203: Use the flower pollination algorithm to optimize and analyze the physiological and ecological parameters to establish a set of parameters applicable to simulating the carbon storage of the mixed forest; S3: Take the mixed forest as the research object and use the improved Biome - BGC model to simulate the carbon storage; S4: Verify the improved Biome - BGC model; S5: Use the extended Fourier amplitude sensitivity test method to analyze the sensitivity of the physiological and ecological parameters; S6: On the basis of selecting highly sensitive parameters, adopt the path analysis method to analyze the positive and negative impacts of highly sensitive parameters on the carbon storage.

2. The carbon storage calculation method of a mixed forest ecosystem according to claim 1, wherein In step 1: The basic geographical data: digital elevation model, slope, aspect, soil sand, clay, and silt contents, short - wave albedo, nitrogen deposition, and nitrogen fixation; The meteorological data: daily maximum temperature, daily minimum temperature, daily average temperature, daily precipitation, daily saturation vapor pressure deficit, daily short - wave radiation flux density, and day length; The physiological and ecological parameters: On the basis of the physiological and ecological parameters defined in the original model, add three parameters to calculate the transfer and litter - fall amounts, including the proportion of evergreen vegetation and deciduous vegetation, the proportion of the transition growth period in the growth season of deciduous vegetation, and the proportion of the litter - fall process in the growth season of deciduous vegetation; The operation history data: Define 10 parameters in the thinning operation management module, including the thinning date, the harvesting rate and transportation rate of each organ of the plant; The verification data: Use the measured data for verification.

3. The carbon storage calculation method of a mixed forest ecosystem according to claim 1, characterized in that In S2011, define a parameter as the proportion of the transition growth period in the growth season of deciduous vegetation to calculate the start and end times of the transition period of deciduous vegetation. Specifically: For deciduous vegetation, the transfer period is described based on the start and end of the growing season. When the sum of the daily average soil temperatures exceeds a defined critical value, leaf unfolding begins; the actual date of leaf unfolding is 15 days earlier than the calculated date, which is the start of the growing season: onset day = m(ST soil ≥T crit ) Act onset_day = onset day -15 where, Tsoil_avg i is the average soil temperature on the i-th day of the year, ST soil is the sum of the daily average soil temperatures when the average soil temperature is greater than 0, T avg is the average of the daily average temperatures during the operating days, T crit is the defined critical value, onset day is the calculated leaf unfolding day, Act onset_day is the actual leaf unfolding day; m is the day when ST soil is greater than or equal to T crit ; If it exceeds July 1st, and the day length is less than 10 hours and 55 minutes; and the soil temperature is lower than the average soil temperature in autumn or the soil temperature is lower than 2°C, all the leaves fall. The actual date when all the leaves fall is 15 days later than this calculated date, and this is the end of the growth season; When one of the following conditions is satisfied, all the leaves fall: offset day = j Actoffset_day = offset day +15 where Daylen j is the day length on the jth day of the year, Tsoil_avg j is the average soil temperature on the jth day of the year, Tsoil avg_aut is the average soil temperature between September and October, offset day is the calculated day when all leaves fall, Actoffset_day is the actual day when all leaves fall, and the growing season is calculated based on the actual leaf emergence day and the day when all leaves fall; ngrowthdays = Actoffset_day - Act onset_day t1 = Act onset_day t2 = Actoffset_day + ngrowthdays × T t_d Among them, ngrowthdays is the number of days in the growing season, t1 and t2 are respectively the start date and end date of the transition period in deciduous vegetation, and T t_d is the proportion of the transition growth period in the growing season of deciduous vegetation.

4. The carbon storage calculation method of a mixed forest ecosystem according to claim 1, characterized in that In S2012, the parameters are defined as the proportions of evergreen vegetation and deciduous vegetation to calculate the daily transfer amounts in different phenological periods of the mixed forest, specifically as follows: Among them, S daily_transfer is the daily transfer amount of the mixed forest, C transfer is the transfer amount of each plant organ in the mixed forest, ndays_E is the remaining transfer days of the evergreen vegetation, ndays_D is the remaining transfer days of the deciduous vegetation, E:D is the ratio of the evergreen vegetation to the deciduous vegetation, and nday represents the day of the year.

5. The carbon storage calculation method of a mixed forest ecosystem according to claim 1, characterized in that S2013: Define the parameter as the proportion of the litterfall process in the growing season of deciduous vegetation, and describe the start and end times of the litterfall process of deciduous vegetation, specifically: t3 = Actoffset_day - (ngrowthdays × LFG d ) + 1 t4 = Actoffset_day Among them, t3 and t4 are respectively the start date and end date of the litterfall process in deciduous vegetation, and LFG d is the proportion of the litterfall process in the growing season of deciduous vegetation; According to the proportion parameters of evergreen vegetation and deciduous vegetation, calculate the daily litterfall of the mixed forest in different phenological periods, specifically as follows: Among them, S daily_litterfall is the daily litterfall amount, Clitterfall_increment_E is the daily litterfall amount of evergreen plants and is a constant ratio throughout the year, Clitterfall_increment_D is the daily litterfall amount of deciduous vegetation, C snnmax is the annual maximum daily carbon, F_turnover is the annual turnover rate, C leaf_froot is the carbon content in leaves or fine roots, litdays D is the remaining litter days of deciduous vegetation, Drate is the linear growth rate, t represents the number of days required to remove all fine roots and fine leaves and the number of iterations in the equation, is the daily litterfall of deciduous vegetation on the t-th day, is the daily litter - fall amount of deciduous vegetation on the (t + 1) - th day.

6. The carbon storage calculation method of a mixed forest ecosystem according to claim 1, characterized in that, In S202, an intermediate cutting management module is added to simulate the impact of intermediate cutting on the carbon storage of mixed forests, specifically: Define the intermediate cutting date, the harvesting rate and transportation rate parameters of each plant organ. The harvesting rate is the proportion of biomass removed during intermediate cutting, and the transportation rate is the proportion that is removed from the site after intermediate cutting and no longer participates in the carbon cycle after transportation is completed. The carbon loss and the amount of carbon entering the next litter pool of each plant organ are calculated according to the intermediate cutting rate and transportation rate, specifically: C all_to_THN = C all × THN C all_to_THN_litr = C all_to_THN × (100 - TRAN) / 100 × litter Among them, THN is the harvesting rate, C all is the carbon content of each organ, C all_to_THN describes the carbon flux generated by thinning in each organ, which is regarded as the carbon loss of each organ. TRAN is the transport rate of each organ, and litter is the proportion of each organ entering the four litter pools, C all_to_THN_litr describes the carbon that remains on-site after thinning and is converted into the corresponding litter components.

7. The carbon storage calculation method of a mixed forest ecosystem according to claim 1, characterized in that, In S203, the flower pollination algorithm is used to optimize and analyze the physiological and ecological parameters to establish a set of parameters suitable for simulating the carbon storage of mixed forests, specifically: Set the maximum number of iterations, the number of individuals in the population, the transfer probability. The objective function is the minimum residual sum between the simulated value and the measured value of carbon storage, and the decision variables are the physiological and ecological parameters to be optimized. Repeatedly adjust the values of the parameters within the feasible range until the objective function reaches the ideal minimum value.

8. The method for calculating the carbon storage of a mixed forest ecosystem according to claim 1, wherein In S3, taking the mixed forest as the research object, the improved Biome-BGC model is used to simulate the carbon storage: In the first step, the spin-up process is used to make the model enter a stable state, and the model outputs a restart file; In the second step, the model uses the restart file as input and starts running forward from this point. The intermediate cutting management module is started in the corresponding area where intermediate cutting management is carried out to simulate the impact of intermediate cutting on carbon storage.

9. The carbon storage calculation method of a mixed forest ecosystem according to claim 1, characterized in that, In S4, the improved Biome-BGC model is verified. The method is: In the first step, the phenology module of the original model is set to evergreen phenology and deciduous phenology for simulation respectively, and then weighted average is carried out according to the proportion of the mixed forest. The simulation results are compared with the simulation results of the improved model adjusted by the phenology module. In the second step, the simulation results of the improved model adjusted by the phenology module are compared with the simulation results of the improved model with both the phenology module adjusted and the intermediate cutting module added; all simulation processes are carried out using the optimized eco-physiological parameters. By calculating R 2 , MAE, and RMSE to evaluate the simulation results of the improved Biome-BGC model, the calculation formulas are as follows: Among them, n represents the number of measured data, and R 2 reflects the fitting degree, and MAE and RMSE reflect the degree of difference. The closer R 2 is to 1, the lower the values of MAE and RMSE, indicating the higher the simulation accuracy.

10. The carbon storage calculation method for a mixed forest ecosystem according to claim 1, characterized in that, In S5, the method for analyzing the sensitivity of physiological and ecological parameters using the extended Fourier amplitude sensitivity test method is as follows: Y = f(X) = f(X1, X2, X3, …, X n ) S i = V i / V Y S iT = S i + S ij + S ijk +…+ S i2…n Among them, Y is the output result of the improved Biome-BGC model, and X i are various physiological and ecological parameters within a given distribution range, and V Y is the total variance of the output, and V i is the variance of a single parameter, and V ij -V 12…n is the variance of the interaction between parameters, and S i is the first-order sensitivity index, representing the direct contribution of the parameter to the total variance of the output result, and S iT is the global sensitivity index, which is the sum of the first-order sensitivity of the parameter and the sensitivity indices of all orders of the interaction between the parameter and other parameters.

Images