A method for estimating dominant tree species bvocs emissions based on remote sensing biomass inversion

Abstract

The present application relates to a kind of dominant tree species BVOCs emission estimation method based on remote sensing biomass inversion, comprising: 1) according to the selection of dominant tree species of forest type in study area;2) for the dominant tree species, selected sample plot in study area and carried out field investigation to obtain the measured parameter related to the above-ground biomass of each dominant tree species;And obtain remote sensing image in study area;3) construct biomass fitting system parameter;4) for each dominant tree species, based on the biomass density of step 3) and remote sensing parameter, with biomass as dependent variable, carry out regression fitting of multiple models;5) establish evaluation index, the accuracy of each model fitting result is evaluated, then the model of optimal accuracy is selected to determine the optimal equation of dominant tree species biomass and vegetation index in study area;6) construct dominant tree species BVOCs emission model based on biomass remote sensing interpretation;And 7) utilize BVOCs emission model to estimate dominant tree species BVOCs emission.

Description

Technical Field

[0001] This invention relates to the field of ecological and environmental technology, and more specifically to a method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion. Background Technology

[0002] Forests are the main body of terrestrial ecosystems, covering approximately 30% of the Earth's land surface, and play a vital role in the carbon cycle, water cycle, and radiative energy exchange of terrestrial ecosystems. Biogenic volatile organic compounds (BVOCs) emitted by forests account for about 70% of the total BVOCs emitted by terrestrial vegetation. The main components of these BVOCs are isoprene, monoterpenes, and oxygenated VOCs, which have strong photochemical activity and exert a significant impact on regional and local air quality and ozone formation.

[0003] The amount of BVOCs emitted by forests is closely related to forest biomass. Existing methods for assessing forest aboveground biomass primarily rely on traditional biomass research methods, namely, obtaining data from sample plots through manual field measurements. For example, forest inventory data is largely based on manual field measurements and primarily measures total biomass. However, current estimates of forest biomass go beyond simply calculating the total amount. There is a need to further subdivide the biomass differences by region, tree age, and tree species to obtain more detailed biomass data. Furthermore, traditional single biomass methods are no longer suitable for large-scale biomass estimation. While remote sensing technology is increasingly advanced, and optical remote sensing imagery is widely used for land change and forest resource monitoring, its rich spectral information effectively reflects vegetation growth. However, cloud cover, missing data bands, and other factors contribute to significant uncertainty in aboveground biomass data.

[0004] Currently, the amount of BVOCs emitted by forests cannot be determined with tree species-level resolution. For example, previous studies using remote sensing interpretation to obtain biomass data for calculating BVOC emissions could only distinguish vegetation types such as broadleaf and coniferous trees, but not tree species. Furthermore, the intensity of volatile organic compound emissions from plants varies by orders of magnitude with different tree species, resulting in significant errors in the estimation results. These issues have been a challenge in previous research and are among the problems that urgently need to be addressed in the field of ecological and environmental research.

[0005] Therefore, new technologies and methods are needed to at least partially solve the problems existing in the aforementioned prior art. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention combines multiple data sources to invert forest aboveground biomass of pure forest tree species in the study area. By evaluating the applicability of the model, it aims to provide a more accurate method for remote sensing estimation of forest aboveground biomass, and improve the biomass differentiation accuracy to the tree species level in the remote sensing interpretation results, so as to achieve high spatiotemporal resolution of BVOCs emission estimation results over a large scale.

[0007] According to one aspect of the present invention, a method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion is provided, comprising:

[0008] 1) Select dominant tree species based on the forest type of the study area;

[0009] 2) For the dominant tree species, select sample plots in the study area and conduct field surveys to obtain measured parameters related to the aboveground biomass of each dominant tree species; and acquire remote sensing images of the study area;

[0010] 3) Construct biomass fitting system parameters, including calculating the biomass density of dominant tree species in the sample plot based on the measured parameters in step 2), and extracting remote sensing parameters related to the remote sensing inversion of forest aboveground biomass based on remote sensing images;

[0011] 4) For each dominant tree species, based on the biomass density and remote sensing parameters from step 3), a multi-model regression fitting was performed with biomass as the dependent variable.

[0012] 5) Establish evaluation indicators to assess the accuracy of the fitting results of each model, and select the model with the best accuracy to determine the biomass of the dominant tree species in the study area; and

[0013] 6) Construct a BVOCs emission model for dominant tree species based on biomass remote sensing interpretation; and

[0014] 7) Use models to estimate BVOC emissions from dominant tree species.

[0015] According to the method of the present invention, the selection of dominant tree species in step 1) includes: black locust, oak, poplar, Chinese pine, larch and birch.

[0016] According to the method of the present invention, the measured parameters in step 2) include tree height, diameter at breast height, crown width, plot area and geographic coordinates; the remote sensing images include Landsat-8OLI satellite data with a spatial resolution of 30m, Sentinel-2 satellite data with a spatial resolution of 10m and ASTER_GDEM_V3 satellite DEM data with a resolution of 30m.

[0017] According to the method of the present invention, the remote sensing image is processed by radiometric calibration, atmospheric correction and spectral reflectance calculation.

[0018] According to the method of the present invention, in step 3), the aboveground biomass density is estimated by using a biomass diameter at breast height (DBH) and tree height regression model; the remote sensing parameters include vegetation index and original band.

[0019] According to the method of the present invention, the vegetation indices include NDVI, RVI, EVI, DVI, MSAVI, SAVI and ARVI, and the original bands are Band 1-Band 7 of Landsat-8OLI and Band 2-Band 8 of Sentinel-2.

[0020] According to the method of the present invention, the multi-model in step 4) includes a multiple linear regression model and a single-factor curve regression model. For example, the single-factor curve regression model includes linear functions, logarithmic functions, quadratic functions, cubic functions, growth functions, inverse functions, S-curve functions, power functions, exponential functions, composite functions, and logistic functions.

[0021] According to the method of the present invention, in step 5), the evaluation index includes the coefficient of determination (R²). 2 The formulas for RMSE and root mean square error are as follows:

[0022]

[0023]

[0024] In the formula, y i This represents the measured value of aboveground biomass. This represents the predicted value of aboveground biomass from the model. This represents the mean of the measured aboveground biomass, and n represents the sample size.

[0025] According to the method of the present invention, step 6 includes obtaining the optimal equation for leaf biomass and vegetation index of the dominant tree species based on the optimal equation for biomass and vegetation index of the dominant tree species using the following formula.

[0026] B 总 =aV+b

[0027]

[0028] In the formula, B 总 V represents aboveground biomass per unit area; a and b are constants; B 叶 D represents the leaf biomass of a tree species. 干 P represents the basic density of the tree trunk, the ratio of the dry weight of the timber to the volume of the green timber. 叶 The proportion of leaf biomass to total biomass; P 干 This refers to the proportion of trunk biomass to total biomass.

[0029] According to the method of the present invention, step 6 includes obtaining a BVOCs emission model based on the optimal equation of dominant tree species leaf biomass and vegetation index using the following formula.

[0030] E ISOP =B 叶 ·ε·γ t ·γ p

[0031] E TMT E OVOC =B 叶 ·ε·γ t

[0032] In the formula: E ISOP E represents the emissions of isoprene; TMT E OVOC ε represents the emission flux of plant monoterpenes and other VOCs; ε represents the VOC emission factor for each forest tree species under standard conditions; B 叶 Tree species leaf biomass; γ t γ p These are dimensionless parameters, namely, the temperature correction factor and the photosynthetically active radiation correction factor.

[0033] This invention combines manual surveys with remote sensing imagery, utilizing multi-source data and multiple models for fitting. In the remote sensing interpretation results, the accuracy of biomass differentiation is improved to the tree species level, achieving high spatiotemporal resolution in BVOCs emission estimation over a large scale, resulting in better fitting and higher accuracy. Furthermore, the method of this invention is widely applicable, thereby improving efficiency and reducing labor costs. In addition, using multi-source remote sensing data simultaneously can reduce interference from objective factors such as cloud cover, achieving scientifically sound results. Attached Figure Description

[0034] The objectives and features of the present invention will become more apparent from the following description taken in conjunction with the accompanying drawings, in which:

[0035] Figure 1 This is a flowchart illustrating the method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion according to an embodiment of the present invention.

[0036] Figure 2 This is a schematic diagram showing the distribution of field survey points in a study area according to an embodiment of the present invention.

[0037] Figure 3 A schematic diagram of the simulated biomass of dominant tree species obtained by the remote sensing biomass inversion-based BVOCs emission estimation method according to an embodiment of the present invention; and

[0038] Figure 4 This is a schematic diagram of the simulation results of BVOCs emissions of dominant tree species in a certain study area according to the remote sensing biomass inversion method for estimating BVOCs emissions of dominant tree species based on an embodiment of the present invention. Detailed Implementation

[0039] The following description, in conjunction with the accompanying drawings, further illustrates the biomass remote sensing interpretation process and method based on plot integration according to the present invention. The following description is merely exemplary and not intended to limit the scope of the invention.

[0040] Figure 1 This is a flowchart illustrating a method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion according to an embodiment of the present invention. As shown in the figure, the method of the embodiment may include the following steps:

[0041] First, the research area is determined, and dominant tree species are selected based on the forest type of the area. The selection of dominant tree species can be achieved through various methods and data, such as the geographical location (latitude and longitude), climate, field surveys, forestry statistics, and the point-to-point conversion function in ArcGIS Excel. For example, dominant tree species in Northeast China include pine, birch, and oak.

[0042] Then, relevant data information within the study area is acquired, including selecting sample plots within the study area for dominant tree species and conducting field surveys to obtain measured parameters related to the aboveground biomass of each dominant tree species. More specifically, for example, one or more sample plots can be selected for each dominant tree species, and the dominant tree species in the sample plots can be monitored. The monitored parameters can include geographic coordinates, tree diameter at breast height (DBH), tree height, crown width, sample plot area, etc. In addition, remote sensing imagery of the study area is acquired. Since trees have a long growth period, remote sensing imagery data during or near the aforementioned field survey period can be acquired without significantly affecting the research and evaluation results. For example, Landsat-8OLI satellite data with a spatial resolution of 30m, Sentinel-2 satellite data with a spatial resolution of 10m, ASTER_GDEM_V3 satellite DEM data with a resolution of 30m, and other suitable remote sensing imagery can be used. To eliminate the adverse effects of cloud cover, missing data stripes, and other factors, remote sensing images can be processed by radiometric calibration, atmospheric correction, and spectral reflectance calculation. These processing techniques are well known in the field and will not be elaborated here. In addition, remote sensing images from multiple different sources can be used simultaneously.

[0043] Next, a biomass fitting system parameter was constructed, including calculating the aboveground biomass density of dominant tree species in the sample plots based on the aforementioned field monitoring data, and extracting remote sensing parameters related to forest aboveground biomass remote sensing inversion from remote sensing images. More specifically, for example, a regression model of single tree species biomass diameter at breast height (DBH) and tree height can be used to calculate the aboveground biomass density of each dominant tree species; spectral reflectance and vegetation indices are important variables in establishing a forest aboveground biomass remote sensing inversion model, from which vegetation indices and original bands can be extracted from remote sensing images, for example, using ENVI software, with DEM as a common parameter. Vegetation indices can include, for example, NDVI, RVI, EVI, DVI, MSAVI, SAVI, and ARVI, etc.

[0044] NDVI stands for Normalized Difference Vegetation Index, and its definition is as follows:

[0045] The difference between the reflectance values ​​in the near-infrared band and the red band of a remotely sensed image is divided by the sum of the two values.

[0046] NDVI = (R NIR -R red ) / (R NIR +R red )

[0047] In the formula, RNIR represents the pixel value (nm) in the infrared band; Rred represents the pixel value (nm) in the red band.

[0048] RVI stands for Ratio Vegetation Index. Because the spectral responses of green plants in the visible red band (R) and near-infrared band (NIR) are significantly different, a simple numerical ratio between the two can adequately express the difference between their reflectivities. Its definition is as follows:

[0049] RVI = R NIR / R red

[0050] In the formula, R NIR Represents the pixel value (nm) in the infrared band; R red This represents the pixel value (nm) in the red light band.

[0051] EVI stands for Perpendicular Vegetation Index. EVI combines the advantages of both the Atmospheric Resistance Vegetation Index and the Soil-Regulating Vegetation Index; its definition is as follows:

[0052]

[0053] In the formula, RNIR Represents the pixel value (nm) in the infrared band; R red R represents the pixel value (nm) in the red light band. blue This represents the pixel value (nm) for the blue band.

[0054] DVI stands for Difference Vegetation Index, which is expressed as the difference between the near-infrared band and the visible light band.

[0055] DVI=R NIR -R red

[0056] In the formula, R NIR Represents the pixel value (nm) in the infrared band; R red This represents the pixel value (nm) in the red light band.

[0057] MSAVI stands for Modified Soil Adjustment Vegetation Index; its definition is as follows:

[0058]

[0059] In the formula, R NIR Represents the pixel value (nm) in the infrared band; R red This represents the pixel value (nm) in the red light band.

[0060] SAVI stands for Soil-Adjusted Vegetation Index, used to account for changes in background optical characteristics and correct for the sensitivity of NDVI to soil background; its definition is as follows:

[0061]

[0062] In the formula, RNIR represents the pixel value (nm) in the infrared band; Rred represents the pixel value (nm) in the red band.

[0063] ARVI stands for Atmospherically Resistant Vegetation Index, which introduces the blue band to counteract the effects of atmospheric transport in the red band; its definition is as follows:

[0064]

[0065] In the formula, RNIR represents the pixel value (nm) of the infrared band; Rred represents the pixel value (nm) of the red band; and Rblue represents the pixel value (nm) of the blue band.

[0066] Subsequently, based on the aforementioned biomass density and remote sensing parameters, a multi-model regression fitting was performed with biomass as the dependent variable. More specifically, methods such as multiple linear regression and single-factor curve regression can be used to regress the remote sensing indices against the measured biomass of each tree species.

[0067] For example, single-factor curve regression can include linear functions, logarithmic functions, quadratic functions, cubic functions, growth functions, inverse functions, S-curve functions, power functions, exponential functions, composite functions, logistic functions, etc. In this invention, forest aboveground biomass data is used as the dependent variable, and remote sensing spectral information, vegetation indices, and DEM elevation data are used as independent variables. A model is constructed through multiple regression analysis to estimate forest aboveground biomass. Calculation formula:

[0068] Y = β0 + β1x1 + β2x2 + ... + β p x p +ε

[0069] In the formula, Y represents the measured value of aboveground biomass (m³). 3 / hm 2 ); x1, x2…x n Let x represent the independent variables affecting biomass. p Cutoff indicates that there are p independent variables, β0, β1, ..., β2. p Let represent the unknown parameters, and ε be the error term.

[0070] If there are n sets of samples, then this linear regression will form a matrix, denoted as (x...). i1 ,x i2 ,…,x ip ,y i ), (i=1,2,…,n), let

[0071]

[0072] Therefore, the matrix form of the multiple linear regression equation is: Y = βX + ε.

[0073] Multiple linear regression involves the relationship between a dependent variable and multiple independent variables. Due to the large number of influencing factors, the regression has high accuracy.

[0074] To verify the accuracy of the simulation results of each model, the coefficient of determination (R²) can be used in this invention. 2 The model results are evaluated using the root mean square error (RMSE) and the coefficient of determination (R²). 2 The formulas for calculating the root mean square error (RMSE) are as follows:

[0075]

[0076] In the formula, y i This represents the measured value of aboveground biomass. This represents the predicted value of aboveground biomass from the model. This represents the mean of the measured aboveground biomass, and n represents the sample size.

[0077] Finally, the optimal aboveground biomass fitting model based on tree species was selected, that is, the model with the best accuracy was selected to determine the biomass of the dominant tree species in the study area.

[0078] Then, a dominant tree species BVOCs emission model based on biomass remote sensing interpretation is constructed, and this model can be used to assess the BVOCs emissions of dominant tree species.

[0079] The construction of the BVOCs emission model for dominant tree species involves using the following formulas to obtain the optimal equations for leaf biomass and vegetation indices of dominant tree species based on the optimal equations for dominant tree species biomass and vegetation indices.

[0080] B 总 =aV+b

[0081]

[0082] In the formula, B 总 V represents aboveground biomass per unit area; a and b are constants; B 叶 D represents the leaf biomass of a tree species. 干 P represents the basic density of the tree trunk, the ratio of the dry weight of the timber to the volume of the green timber. 叶 The proportion of leaf biomass to total biomass; P 干 This refers to the proportion of trunk biomass to total biomass.

[0083] The construction of the dominant tree species BVOCs emission model also includes using the following formula to obtain the BVOCs emission model based on the optimal equation of leaf biomass and vegetation index of the dominant tree species.

[0084] E ISOP =B 叶 ·ε·γ t ·γ p

[0085] E TMT E OVOC =B 叶 ·ε·γ t

[0086] In the formula: E ISOP E represents the emissions of isoprene; TMT E OVOCε represents the emission flux of plant monoterpenes and other VOCs; ε represents the VOC emission factor for each forest tree species under standard conditions; B 叶 Tree species leaf biomass; γ t γ p These are dimensionless parameters, namely, the temperature correction factor and the photosynthetically active radiation correction factor.

[0087] Example: The following example, using a region in North China, further illustrates the method of biomass of dominant tree species in a study area based on remote sensing interpretation.

[0088] First, the survey and sampling phase included field surveys and satellite image extraction: Using forest resource inventory data and ArcGIS Excel's point-to-point function, the main dominant tree species in the study area were selected based on their distribution and scale. These included: black locust, oak, poplar, Chinese pine, larch, and birch. Field surveys were conducted from July to September 2021, including: coordinates, diameter at breast height (DBH) per tree, tree height per tree, crown width, and plot area. For satellite image extraction, Landsat-8OLI satellite data with a spatial resolution of 30m (after radiometric calibration, atmospheric correction, and spectral reflectance calculations), Sentinel-2 satellite data with a spatial resolution of 10m, and ASTER_GDEM_V3 satellite DEM data with a resolution of 30m were selected.

[0089] Second, based on the above-mentioned survey and sampling results, a biomass fitting system parameter was constructed, specifically including: 1) Calculating the measured aboveground biomass density: For example, selecting a biomass, diameter at breast height (DBH), and tree height regression model for a single tree species in the study area and North China from standards and literature, and combining the biomass results with the sample plot area (20×20m) to calculate the aboveground biomass density. 2) Extracting remote sensing image fitting parameters: Spectral reflectance and vegetation indices are important variables for establishing a forest aboveground biomass remote sensing inversion model. Therefore, this invention selects three parts: original bands, vegetation indices, and DEM data, totaling 15 bands. Among the original bands, for Landsat-8OLI, Bands 1-7 were selected, and for Sentinel-2 original bands, Bands 2-8 were extracted; the vegetation indices were extracted using Landsat-8OLI and Sentinel-2 respectively: NDVI, RVI, EVI, DVI, MSAVI, SAVI, and ARVI. The specific calculation methods for each vegetation index extraction are explained in detail below.

[0090] 1) Calculate the measured aboveground biomass density.

[0091] Using the diameter at breast height (DBH) and tree height information from the sample plots, a DBH-tree height regression model was used in conjunction with the forest area to calculate the W value of tree trunks per hectare. s Branches Wb Leaves W l Tree roots W r Biomass, and then further calculate the total biomass (m). 3 / hm 2 The results are shown in Table 1 below:

[0092] Table 1 Standard Equation of Regression Model for Biomass, Diameter at Breast Height, and Tree Height

[0093]

[0094] 2) Extraction of parameters for remote sensing image fitting

[0095] Using the relevant remote sensing images obtained, and based on the definitions of the above indices, the Normalized Difference Vegetation Index (NDVI), Ratio Vegetation Index (RVI), Enhanced Vegetation Index (EVI), Difference Vegetation Index (DVI), Adjusted Soil-Regulated Vegetation Index (MSAVI), Soil-Regulated Vegetation Index (SAVI), and Atmospheric Resistance Vegetation Index (ARVI) were calculated.

[0096] Third, after constructing the parameters for the biomass fitting system, the extracted remote sensing indices were regressed against the measured biomass. The regression methods mainly included multiple linear regression and single-factor curve regression. The specific calculation methods for these two approaches are detailed below.

[0097] 1) Multiple linear regression model

[0098] Multiple linear regression models are widely used in forest biomass estimation. Typically, forest aboveground biomass data is used as the dependent variable, and remote sensing spectral information, vegetation indices, etc., are used as independent variables. A model is constructed through multiple regression analysis to estimate forest aboveground biomass. Calculation formula:

[0099] Y = Xβ + ε

[0100] In the formula, Y represents the measured value of aboveground biomass (m3 / hm2); X represents the independent variable factor affecting biomass; β is an unknown parameter; and ε is the error term.

[0101] 2) Single-factor curve regression model

[0102] Based on SPSS's regression analysis function, 15 extracted remote sensing indices were selected as independent variables, with biomass as the dependent variable, and model fitting was performed for each. The regression models included: linear function, logarithmic function, quadratic function, cubic function, growth function, inverse function, S-curve function, power function, exponential function, composite function, and logistic function. The model expressions selected for linear and curve fitting are shown in Table 2 below:

[0103] Table 2 General Expression of Fitting Model

[0104] Model General expression linear y = a + bx logarithm y = a + bln(x) Quadratic terms <![CDATA[y=a+bx+cx 2 ]]> cube <![CDATA[y=a+bx+cx 2 +dx 3 ]]> increase <![CDATA[y=e a+bx ]]> Inverse model y = a + b / x S-curve <![CDATA[y=e a+b / x ]]> power <![CDATA[y=ax b ]]> Exponential function <![CDATA[y=a x ]]> Composite function y = abx Logistic function <![CDATA[logit(y)=c+b1x1+b2x2]]>

[0105] Fourth, establish evaluation indicators to assess the accuracy of the fitting results of each model; in order to make full use of the sample and improve the reliability of the model, the coefficient of determination (R²) is adopted. 2 The model is evaluated using the root mean square error (RMSE) and the mean square error (RMSE). The calculation method is as follows:

[0106]

[0107] In the formula, y i This represents the measured value of aboveground biomass. This represents the predicted value of aboveground biomass from the model. This represents the mean of the measured aboveground biomass, and n represents the sample size.

[0108] The data for each indicator were categorized, and the six selected tree species were fitted to two different satellite data sets using both single-factor curve fitting and multiple linear regression. The evaluation results obtained from fitting different satellite data sets and methods were compared and analyzed, and the optimal aboveground biomass fitting equation based on the tree species was ultimately selected.

[0109] The results are shown in Table 3.

[0110] Table 3. Accuracy of Biomass Inversion Results

[0111]

[0112]

[0113] The results showed that, for the biomass of the six dominant tree species in the study area, the results of the multiple linear regression model were generally higher than those of the single-factor curve regression model. More specifically, for Robinia pseudoacacia, the inversion results from 13 bands of the Landsat-8 satellite were the best, with a fitting R0. 2 The value is 0.304; for oak trees, the retrieval results for 14 bands from the Sentinel-2 satellite are the best, with a fitting R value of 0.304. 2 The value is 0.379; for poplar, the retrieval results for 14 bands from the Sentinel-2 satellite are the best, with a fitting R value of 0.379. 2 The value is 0.407; for Pinus tabuliformis, the inversion results of 15 bands from the Landsat-8 satellite are the best, with a fitting R value of 0.407. 2 The value is 0.502; for birch, the inversion results for 11 bands from the Sentinel-2 satellite are the best, with a fitting R value of 0.502. 2 The value is 0.533; for larch, the retrieval results for 15 bands from the Landsat-8 satellite are the best, with a fitting R value of 0.533. 2 It is 0.594.

[0114] Fifth, based on the fitting results, the optimal equations for the biomass of dominant tree species and vegetation index in the study area were obtained. Specific results are shown in Table 4.

[0115] Table 4 Optimal equations for biomass and vegetation index of dominant tree species in the study area

[0116]

[0117] The biomass of dominant tree species in the entire study area was simulated using the optimal equations for the biomass and vegetation index of the dominant tree species in the study area. The results are as follows: Figure 3 As shown.

[0118] Sixth, the biomass equation data is used to obtain the leaf biomass equation as the BVOCs emission calculation equation, which includes two parts: parameter acquisition and model construction.

[0119] 1) Parameter Acquisition

[0120] The calculation of BVOCs emissions requires the volume V as an intermediate parameter to calculate leaf biomass B. 叶 The volumetric volume was calculated by back-calculating using the conversion factor continuous function method based on the biomass-volume conversion factor parameters provided in the "Assessment of Biomass and Carbon Storage in China's Forest Vegetation". Following the "Technical Guidelines for Compiling Dynamic High Spatiotemporal Resolution Natural Source VOCs Emission Inventory", the obtained volumetric volume data was substituted into the leaf biomass equation required for calculating BVOCs. The parameter calculation method is as follows:

[0121] B 总 =aV+b

[0122]

[0123] In the formula, B 总 The aboveground biomass per unit area (calculated using Table 4) is expressed in t / hm². 2 V represents the volume per unit area, in meters. 3 a and b are both constants; B 叶 D represents the leaf biomass of the tree species in g; 干 The basic density of the tree trunk (the ratio of the mass of oven-dry wood to the volume of green wood), in g·m³. -3 ;P 叶 The percentage of leaf biomass in total biomass, %; P 干 The percentage of trunk biomass to total biomass.

[0124] Other parameters in the formula are shown in Table 5 and are determined according to the "Assessment of Forest Vegetation Biomass and Carbon Storage in China" and the "Technical Guidelines for Compiling Dynamic High Spatiotemporal Resolution Natural Source VOCs Emission Inventory".

[0125] Table 5 Summary of parameters for calculating leaf biomass

[0126] species a b <![CDATA[D 干 (g·m -3 )]]> <![CDATA[P 叶 (%)]]> <![CDATA[P 干 (%)]]> locust 0.7564 8.3103 0.598 5.55 53.57 oak 1.1453 8.5473 0.676 5.55 53.57 poplar 0.4754 30.6034 0.378 5.55 53.57 Pinus tabuliformis 0.7554 5.0928 0.360 9.48 55.99 Birch 1.0687 10.2370 0.541 5.55 53.57 larch 0.6096 33.8060 0.490 3.47 63.11

[0127] Substituting the obtained parameters into the biomass equation in step five yields the optimal equations for leaf biomass and vegetation indices of the dominant tree species in the study area. See Table 6:

[0128] Table 6. Optimal equations for biomass and vegetation index of dominant tree species in the study area.

[0129]

[0130] 2) Construction of BVOCs emission model

[0131] Natural source VOCs emitted by forest trees can be classified into isoprene, monoterpenes, and other VOCs. Based on the G93 model and considering the effects of light and temperature, the specific calculation method for BVOCs is as follows:

[0132] E ISOP =B 叶 ·ε·γ t ·γ p (Isoprene)

[0133] E TMT E OVOC =B 叶 ·ε·γ t (Monoterpenes, other VOCs)

[0134] In the formula: E ISOP ε represents the emissions of isoprene; ETMT and EOVOC represent the emissions of plant monoterpenes and other VOCs fluxes; ε is the VOC emission factor for each forest tree species under standard conditions; B 叶 γ represents the leaf biomass of each tree species; γt and γp are dimensionless parameters, representing the temperature correction factor and the photosynthetically active radiation correction factor, respectively.

[0135] Emission factor parameters were constructed: Since the main source of BVOCs emissions is tree leaves, the standard was primarily based on leaf levels. The standard emission factors for the six dominant tree species selected in this study area are shown in Table 7.

[0136] Table 7 Emission rates of dominant tree species

[0137]

[0138] Constructing photo-temperature correction factor parameters: Among BVOCs emissions, isoprene is affected by the combined effects of light and temperature, while monoterpenes and other VOCs are mainly affected by temperature.

[0139] For isoprene, the optical correction factor is calculated as follows:

[0140]

[0141] In the formula: L is the photosynthetically active radiation flux (PAR), μmol·m -2 ·s -1 ; α is an empirical constant, 0.0027; γ L1 It is an empirical constant, 1.006.

[0142] Photosynthetically active radiation (PAR) refers to the portion of solar radiation that can be used by green plants for photosynthesis. The temperature correction factor is calculated as follows:

[0143]

[0144] In the formula: T is the leaf temperature, in K; air temperature can be used instead of leaf temperature; T S The leaf temperature is under standard conditions, with a reference value of 303 K; R is a constant with a value of 8.314 J·K. -1 ·mol -1 ;γ t1 γ t2 All are constants, with values ​​of 95000 J·mol⁻¹. -1 230000 J·mol -1 ;T M It is 314K.

[0145] The specific calculation method for the temperature correction factor of monoterpenes is as follows:

[0146] γ T =exp[β(TT) S )]

[0147] In the formula; β is an empirical parameter, 0.09·K -1 T represents leaf temperature in Kelvin; air temperature is used instead of leaf temperature. S The leaf temperature is 303 K under standard conditions.

[0148] The annual photothermal correction factors for isoprene and monoterpenes, calculated using the above method from daily and hourly meteorological data for 2021, are shown in the table below:

[0149] Table 8 Monthly Data on Light Temperature Correction Factor

[0150] month <![CDATA[Isoprene γ p > <![CDATA[Isoprene γ t > <![CDATA[Monoterpene γ t > January 0.29966538 0.00856741 0.04406756 February 0.34221634 0.03154842 0.09043826 March 0.37259399 0.06670856 0.1500198 April 0.44562446 0.13565337 0.24266123 May 0.48282524 0.2750615 0.38951422 June 0.47576752 0.50986774 0.60983114 July 0.43670529 0.54593989 0.647625 August 0.44181638 0.46258591 0.57212226 September 0.36766358 0.29611673 0.41856063 October 0.3442075 0.09878455 0.1973209 November 0.29109121 0.03658805 0.10542106 December 0.27973793 0.01686943 0.06501561

[0151] By combining research data with established tree species biomass inversion models, the leaf biomass of each tree species in the general emission estimation model of BVOCs was transformed into a mathematical formula with BVOCs as the dependent variable and the remote sensing index as the dependent variable, using the established remote sensing index-leaf biomass model for each tree species. The transformed BVOCs estimation model is shown in Table 9:

[0152] The BVOCs emissions of dominant tree species in the entire study area were simulated using the BVOCs estimation model shown in Table 9 below. The results are as follows: Figure 4 As shown.

[0153] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the device and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

[0154] Table 9 BVOCs Emission Model

[0155]

Claims

1. A method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion, comprising: 1) Select dominant tree species based on the forest type of the study area; 2) For the dominant tree species, select sample plots in the study area and conduct field surveys to obtain measured parameters related to the aboveground biomass of each dominant tree species, as well as obtain remote sensing images of the study area; 3) Construct biomass fitting system parameters, including calculating the biomass density of dominant tree species in the sample plot based on the measured parameters in step 2), and extracting remote sensing parameters related to the remote sensing inversion of forest aboveground biomass based on remote sensing images; 4) For each dominant tree species, based on the biomass density and remote sensing parameters from step 3), a multi-model regression fitting was performed with biomass as the dependent variable. 5) Establish evaluation indicators to evaluate the accuracy of the fitting results of each model, and then select the model with the best accuracy to determine the optimal equation for the biomass of dominant tree species and vegetation index in the study area. 6) Construct a BVOCs emission model for dominant tree species based on biomass remote sensing interpretation; as well as 7) Estimate BVOC emissions from dominant tree species using BVOC emission models; The dominant tree species in step 1) include: black locust, oak, poplar, Chinese pine, larch, and birch; The BVOCs emission model for black locust is as follows: E ISOP = 31.87 [0.082 (91.522*NDVI + 79.124*RVI + 2248.909*DVI - 1.011*DEM + 1748.647*SAVI - 325.7*MSAVI + 1812.743*Band1 + 2885.464*Band2 - 933.534*Band3 - 244.187*Band4 - 1202.848*Band5 + 1122.035*Band6 - 812.057*Band7 - 2106.168 3 )- 0.681] * gamma t * gamma p ; E TMT ,E OVOC = 0.58 [0.082 (91.522*NDVI + 79.124*RVI + 2248.909*DVI - 1.011*DEM + 1748.647*SAVI - 325.7*MSAVI + 1812.743*Bandl + 2885.464*Band2 - 933.534*Band3 - 244.187*Band4 - 1202.848*Band5 + 1122.035*Band6 - 812.057*Band7 - 2106.168) - 0.681] · γ t ; The BVOCs emission model for oak trees is as follows: E ISOP = 110.40 [0.061 (= -2.241e 03 NDVI + 5.524 RVI + 4.797e 02 EVI + 1.219e 03 ARVI - 1.119e -03 DEM - 3.064 * e 03 SAVI + 4.91e 02 MSAVI + 2.082e 03 Band1 - 5.033e 02 Band2 + 3.797e 03 Band3 - 5.796e 03 Band4 - 5.653e 02 Bnad5 + 9.069e 02 Band6 - 6.497e Band7 + 2.579e 03 )- 0.523] · γ t · γ p E TMT ,E OVOC = 3.10 [0.061 (= -2.241e 03 NDVI + 5.524 RVI + 4.797e 02 EVI + 1.219e 03 ARVI - 1.119e -03 DEM - 3.064*e 03 SAVI + 4.91e 02 MSAVI + 2.082e 03 Band1 - 5.033e 02 Band2 + 3.797e 03 Band3 - 5.796e 03 Band4 - 5.653e 02 Bnad5 + 9.069e 02 Band6 - 6.497e Band7 + 2.579e 03 )- 0.523] · γ t ; The BVOCs emission model for poplar trees is as follows: E ISOP = 90.86 [0.082 (83.809 NDVI + 1.346 RVI - 84.822 EVI - 0.004 DEM - 274.446 SAVI - 135.874 MSAVI - 296.488 Band2 - 137.982 Band3 - 477.083 Band4 - 95.680 Band5 - 93.906 Band6 + 78.367 Band7 + 694.454 Band8 + 23.642) - 2.521] · γ t · γ p E TMT ,E OVOC = 4.05 [0.082 (83.809 NDVI + 1.346 RVI - 84.822 EVI - 0.004 DEM - 274.446 SAVI - 135.874 MSAVI - 296.488 Band2 - 137.982 Band3 - 477.083 Band4 - 95.680 Band5 - 93.906 Band6 + 78.367 Band7 + 694.454 Band8 + 23.642) - 2.521] · γ t ; The BVOCs emission model of Pinus tabuliformis is as follows: E ISOP = 0.54 [0.081 (-5.043eNDVI + 5.114eRVI - 4.474e 02 EVI + 3.640e 02 DVI + 2.898e -02 DEM - 7.872e 02 ARVI + 4.255e 02 SAVI + 4.052e 02 MSAVI + 1.542e 03 Band1 - 3.084e 02 Band2 - 2.034e 02 Band3 - 1.044e 02 Band4 - 2.618e 02 Band5 + 8.575e Band6 + 4.726e 02 Band7 - 1.090e 02 )- 0.411] · γ t · γ p E TMT ,E OVOC = 9.46 [0.081 (-5.043eNDVI + 5.114eRVI - 4.474e 02 DVI + 2.898e 02 DVI + 2.898e - 02 DEM - 7.872e 02 ARVI + 4.255e 02 SAVI + 4.052e 02 MSAVI + 1.542e 03 Band1 - 3.084e 02 Band2 - 2.034e 02 Band3 - 1.044e 02 Band4 - 2.618e 02 Band5 + 8.575eBand6 + 4.726e 02 Band7 - 1.090e 02 )- 0.411] · γ t ; The BVOCs emission model for birch trees is as follows: E ISOP = 0.22 [0.052 (1195.375 NDVI + 877.883 RVI - 4017.468 MSAVI + 1321.495 SAVI + 3157.581 Bandl - 4479.349 Band2 - 1705.667 Band3 + 5443.259 Band4 + 1150.283 Band5 - 2105.098 Band6 + 0.069 Band7 - 2917.283) - 0.537] · γ t · γ p E TMT ,E OVOC <0.15[0.052(1195.375NDVI+877.883RVI-4017.468MSAVI+1321.495SAVI+3157.581Band1-4479.349Band2-1705.667Band3+5443.259Band4+1150.283Band5-2105.098Band6+0.069Band7-2917.283)-0.537]·γ t ; The BVOCs emission model for larch is as follows: E ISOP =0.07[0.044(32.907NDVI-27.928RVI+85.452EVI+402.372DVI-326.989MSAVI+402.033ARVI-191.334SAVI-1516.31Band1+1712.918Band2+213.868Band3-683.286Band4-201.329Band5+271.641Band6-0.464Bnad7+0.051Band8+277.208)-1.494]·γ t ·γ p E TMT ,E OVOC <0.05[0.044(32.907NDVI-27.928RVI+85.452EVI+402.372DVI-326.989MSAVI+402.033ARVI-191.334SAVI-1516.31Band1+1712.918Band2+213.868Band3-683.286Band4-201.329Band5+271.641Band6-0.464Bnad7+0.051Band8+277.208)-1.494]·γ t ; Among them, E ISOP E represents the emissions of isoprene; TMT E OVOC For plant monoterpenes emissions and other VOCs emission fluxes; γ t γ p These are the temperature correction factor and the photosynthetically active radiation correction factor, respectively; NDVI, RVI, EVI, DVI, MSAVI, SAVI, and ARVI are common vegetation indices of Landsat-8OLI and Sentinel-2 satellite imagery; Band 2-Band 8 in the oak and poplar models are the original bands of Sentinel-2; Band 1-Band 7 in the black locust, pine, larch, and birch models are the original bands of Landsat-8OLI.

2. The method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion according to claim 1, wherein, The measured parameters in step 2) include tree height, diameter at breast height (DBH), crown width, plot area, and geographic coordinates; the remote sensing images include Landsat-8OLI satellite data with a spatial resolution of 30m, Sentinel-2 satellite data with a spatial resolution of 10m, and ASTER_GDEM_V3 satellite DEM data with a resolution of 30m.

3. The method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion according to claim 2, wherein, In step 3), the aboveground biomass density is estimated using the biomass diameter at breast height (DBH) and tree height regression model method; the remote sensing parameters include vegetation index and original band.

4. The method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion according to claim 1, wherein, The remote sensing image described in step 2) has undergone radiometric calibration, atmospheric correction, and spectral reflectance calculation.

5. The method for estimating BVOCs emissions of dominant tree species based on remote sensing biomass inversion according to claim 1, wherein, The multiple models mentioned in step 4) include multiple linear regression models and single-factor curve regression models.

6. The method for estimating BVOCs emissions of dominant tree species based on remote sensing biomass inversion according to claim 5, wherein, Step 6 includes using the following formula to obtain the optimal equation for leaf biomass and vegetation index of the dominant tree species, based on the optimal equation for biomass and vegetation index of the dominant tree species. B 总 =aV+b In the formula, B 总 V represents aboveground biomass per unit area; a and b are constants; B 叶 D represents the leaf biomass of a tree species. 干 P represents the basic density of the tree trunk, the ratio of the dry weight of the timber to the volume of the green timber. 叶 The proportion of leaf biomass to total biomass; P 干 This refers to the proportion of trunk biomass to total biomass.

7. The method for estimating BVOCs emissions of dominant tree species based on remote sensing biomass inversion according to claim 6, wherein, Step 6 includes obtaining the BVOCs emission model using the following formula, based on the optimal equation of dominant tree species leaf biomass and vegetation index. E ISOP =B 叶 ·e·c t ·c p E TMT ,E OVOC =B 叶 ·e·c t In the formula: ε is the VOC emission factor of each forest tree species under standard conditions; B 叶 This refers to the leaf biomass of tree species.

8. The method for estimating BVOCs emissions from dominant tree species based on remote sensing biomass inversion according to claim 1, wherein, In step 5), the evaluation indicators include the coefficient of determination (R²). 2 The formulas for RMSE and root mean square error are as follows: In the formula, y i This represents the measured value of aboveground biomass. This represents the predicted value of aboveground biomass from the model. This represents the mean of the measured aboveground biomass, and n represents the sample size.

Images