A carbon sink rapid monitoring method and device based on remote sensing

Abstract

The present application discloses a carbon sink rapid monitoring method and device based on remote sensing, which relates to the field of carbon sink monitoring technology. The main technical scheme is: obtaining multi-source remote sensing data corresponding to the target area; determining the measured pixel points in the medium-low resolution vegetation net primary productivity remote sensing product based on the medium-high resolution land classification remote sensing product; calculating the vegetation index of each pixel point based on the band data corresponding to the cloud-free medium-high resolution remote sensing image; inverting the medium-low resolution vegetation net primary productivity of each measured pixel point and the vegetation index of each measured pixel point to obtain the medium-high resolution vegetation net primary productivity of each pixel point in the cloud-free medium-high resolution remote sensing image; calculating the carbon sink of the medium-high resolution vegetation net primary productivity of each pixel point in the cloud-free medium-high resolution remote sensing image based on the medium-high resolution forest age remote sensing product, and obtaining the carbon sink amount of each pixel point in the cloud-free medium-high resolution remote sensing image.

Description

Technical Field

[0001] The present application relates to the technical field of carbon sink monitoring, and in particular to a method and device for rapid carbon sink monitoring based on remote sensing. Background Art

[0002] As one of the largest terrestrial carbon pools on Earth, the net primary productivity (NPP) of vegetation is a key indicator for measuring its carbon sequestration capacity, which directly affects the dynamic balance of regional and even global carbon cycles. Rapid and accurate monitoring of vegetation carbon sinks using remote sensing technology has become a research hotspot in recent years.

[0003] At present, the traditional remote sensing technology to monitor net primary productivity mainly relies on low- and medium-resolution remote sensing images to provide global net primary productivity monitoring. However, the spatial resolution of low- and medium-resolution remote sensing images is low, which cannot meet the needs of high-precision carbon sink monitoring in certain specific areas, and there is heterogeneity between different regions, which makes it difficult to ensure the efficiency and accuracy of carbon sink monitoring, which is not conducive to the research and management of carbon cycle. Summary of the invention

[0004] In view of the above problems, the present application provides a remote sensing-based carbon sink rapid monitoring method and device, the main purpose of which is to improve the efficiency and accuracy of carbon sink monitoring.

[0005] In order to solve the above technical problems, this application proposes the following solutions:

[0006] In a first aspect, the present application provides a method for rapid carbon sink monitoring based on remote sensing, the method comprising:

[0007] Acquire multi-source remote sensing data corresponding to the target area, wherein the multi-source remote sensing data includes cloud-free medium-high resolution remote sensing images, medium-low resolution vegetation net primary productivity remote sensing products, medium-high resolution land classification remote sensing products, and medium-high resolution forest age remote sensing products;

[0008] Determine the quasi-measured pixel points in the medium- and low-resolution vegetation net primary productivity remote sensing product based on the medium- and high-resolution land classification remote sensing product, and extract the medium- and low-resolution vegetation net primary productivity of each of the quasi-measured pixel points, wherein the quasi-measured pixel points are used to characterize the pixel points with uniform vegetation in the medium- and low-resolution vegetation net primary productivity remote sensing product;

[0009] Calculating the vegetation index of each pixel point based on the band data corresponding to the cloud-free high-resolution remote sensing image, and extracting the vegetation index corresponding to each of the measured pixel points;

[0010] Using the low- and medium-resolution net primary productivity of vegetation of each of the measured pixel points and the vegetation index of each of the measured pixel points to perform inversion, the high- and medium-resolution net primary productivity of vegetation of each pixel point in the cloud-free high- and medium-resolution remote sensing image is obtained;

[0011] The carbon sink amount of each pixel point in the cloud-free medium and high resolution remote sensing image is obtained by calculating the carbon sink of the medium and high resolution vegetation net primary productivity of each pixel point in the cloud-free medium and high resolution remote sensing image based on the medium and high resolution forest age remote sensing product.

[0012] In a second aspect, the present application provides a carbon sink rapid monitoring device based on remote sensing, the device comprising:

[0013] A first acquisition unit is used to acquire multi-source remote sensing data corresponding to the target area, wherein the multi-source remote sensing data includes cloud-free medium-high resolution remote sensing images, medium-low resolution vegetation net primary productivity remote sensing products, medium-high resolution land classification remote sensing products, and medium-high resolution forest age remote sensing products;

[0014] A determination unit, used to determine the quasi-measured pixel points in the medium- and low-resolution vegetation net primary productivity remote sensing product based on the medium- and high-resolution land classification remote sensing product of the first acquisition unit, and extract the medium- and low-resolution vegetation net primary productivity of each of the quasi-measured pixel points, wherein the quasi-measured pixel points are used to characterize the pixel points with uniform vegetation in the medium- and low-resolution vegetation net primary productivity remote sensing product;

[0015] A calculation unit, used to calculate the vegetation index of each pixel point based on the band data corresponding to the cloud-free high-resolution remote sensing image of the first acquisition unit, and extract the vegetation index corresponding to each of the quasi-measured pixel points;

[0016] An inversion unit, used to perform inversion using the medium- and low-resolution net primary productivity of vegetation of each of the quasi-measured pixel points of the determination unit and the vegetation index of each of the quasi-measured pixel points of the calculation unit, to obtain the medium- and high-resolution net primary productivity of vegetation of each pixel point in the cloud-free medium- and high-resolution remote sensing image;

[0017] A processing unit is used to calculate the carbon sink of the medium-high resolution vegetation net primary productivity of each pixel point in the cloud-free medium-high resolution remote sensing image of the inversion unit based on the medium-high resolution forest age remote sensing product, so as to obtain the carbon sink amount of each pixel point in the cloud-free medium-high resolution remote sensing image.

[0018] In order to achieve the above-mentioned purpose, according to the third aspect of the present application, a storage medium is provided, which includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the remote sensing-based carbon sink rapid monitoring method of the above-mentioned first aspect.

[0019] In order to achieve the above-mentioned purpose, according to the fourth aspect of the present application, a processor is provided, wherein the processor is used to run a program, wherein when the program is run, the remote sensing-based carbon sink rapid monitoring method of the above-mentioned first aspect is executed.

[0020] By means of the above technical scheme, the present application provides a remote sensing-based rapid carbon sink monitoring method and device. When it is necessary to monitor the vegetation carbon sink, first, obtain the multi-source remote sensing data corresponding to the target area, the multi-source remote sensing data includes cloud-free medium-high resolution remote sensing images, medium-low resolution vegetation net primary productivity remote sensing products, medium-high resolution land classification remote sensing products and medium-high resolution forest age remote sensing products. Then, based on the medium-high resolution land classification remote sensing products, determine the quasi-measured pixel points in the medium-low resolution vegetation net primary productivity remote sensing products, and extract the medium-low resolution vegetation net primary productivity of each quasi-measured pixel point. The quasi-measured pixel points are used to characterize the medium-low resolution vegetation net primary productivity. The pixels with uniform vegetation in the remote sensing products are used. At the same time, the vegetation index of each pixel is calculated based on the band data corresponding to the cloud-free medium and high-resolution remote sensing images, and the vegetation index corresponding to the measured pixel points of each class is extracted. Then, the medium and low-resolution vegetation net primary productivity of each class of measured pixel points and the vegetation index of each class of measured pixel points are used for inversion to obtain the medium and high-resolution vegetation net primary productivity of each pixel point in the cloud-free medium and high-resolution remote sensing images. Finally, based on the medium and high-resolution forest age remote sensing products, the carbon sink of the medium and high-resolution vegetation net primary productivity of each pixel point in the cloud-free medium and high-resolution remote sensing images is calculated to obtain the carbon sink amount of each pixel point in the cloud-free medium and high-resolution remote sensing images. Through the technical solution provided in this application, more detailed vegetation structure and growth status information can be obtained from cloud-free medium and high-resolution remote sensing images, and quasi-measured pixel points that effectively reflect the vegetation carbon sink characteristics of the target area can be selected from medium and low-resolution vegetation net primary productivity remote sensing products and medium and high-resolution land classification remote sensing products. The medium and high-resolution vegetation net primary productivity can be inverted using the vegetation index obtained from the cloud-free medium and high-resolution remote sensing images, and the accurate medium and high-resolution vegetation net primary productivity can be estimated. The carbon sink of the medium and high-resolution vegetation net primary productivity can then be calculated based on the medium and high-resolution forest age remote sensing products. The impact of heterogeneity on carbon sink monitoring is comprehensively considered, and high-precision, rapid, and comprehensive monitoring of vegetation carbon sinks is achieved, thereby effectively improving the efficiency and accuracy of carbon sink monitoring, which is beneficial to the research and management of the carbon cycle.

[0021] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0023] Figure 1 A flow chart of a carbon sink rapid monitoring method based on remote sensing provided in an embodiment of the present application is shown;

[0024] Figure 2 Another flow chart of a carbon sink rapid monitoring method based on remote sensing provided in an embodiment of the present application is shown;

[0025] Figure 3 A block diagram of a carbon sink rapid monitoring device based on remote sensing provided in an embodiment of the present application is shown;

[0026] Figure 4 A block diagram showing the composition of another carbon sink rapid monitoring device based on remote sensing provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0027] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0028] At present, the traditional remote sensing technology to monitor net primary productivity mainly relies on low- and medium-resolution remote sensing images to provide global net primary productivity monitoring. However, the spatial resolution of low- and medium-resolution remote sensing images is low, which cannot meet the needs of high-precision carbon sink monitoring in certain specific areas, and there is heterogeneity between different regions, which leads to inaccurate monitoring of carbon sinks, which is not conducive to the research and management of carbon cycle.

[0029] The inventors have found that in order to achieve high-precision carbon sink monitoring and overcome heterogeneity between regions, a target area can be used as the basis, and the multi-source remote sensing data corresponding to the target area can be used to select quasi-measured pixel points with uniform vegetation. The net primary productivity of vegetation at medium and low resolutions and the corresponding vegetation index at medium and high resolutions can be used to invert the net primary productivity of vegetation at medium and high resolutions, and the carbon sink conversion can be corrected to finally obtain the carbon sink. In this way, the impact of heterogeneity on carbon sink monitoring is comprehensively considered, and high-precision, rapid, and comprehensive monitoring of vegetation carbon sinks is achieved, thereby effectively improving the efficiency and accuracy of carbon sink monitoring, which is conducive to the research and management of carbon cycles.

[0030] To this end, the present application embodiment provides a carbon sink rapid monitoring method based on remote sensing, through which the efficiency and accuracy of carbon sink monitoring can be improved. The specific execution steps are as follows: Figure 1 As shown, including:

[0031] 101. Obtain multi-source remote sensing data corresponding to the target area.

[0032] In this embodiment, the target area is the area where the carbon sink needs to be monitored, and the specific source of the multi-source remote sensing data of the target area can be optical remote sensing satellites (such as Landsat, MODIS, Sentinel, etc.), drone aerial photography data, etc. It is only necessary to ensure that the data covers the target area and meets the requirements of resolution, time span, etc. After obtaining the multi-source remote sensing data, it can be pre-processed by radiation correction, atmospheric correction, geographic registration, etc. to ensure the quality and consistency of the multi-source remote sensing data.

[0033] Multi-source remote sensing data include cloud-free medium-high resolution remote sensing images, medium-low resolution vegetation net primary productivity remote sensing products, medium-high resolution land classification remote sensing products, and medium-high resolution forest age remote sensing products. Among them, cloud-free medium-high resolution remote sensing images refer to remote sensing images with high spatial resolution obtained under cloud-free conditions in the target area, which can provide detailed surface information of the target area, such as satellite data such as Landsat8 and Sentinel-2. Medium-low resolution vegetation net primary productivity remote sensing products are used to provide estimates of medium-low resolution vegetation net primary productivity in the target area, usually from medium-low resolution sensors such as MODIS. Medium-high resolution land classification remote sensing products are used to provide detailed land cover classification information in the target area, including different types of vegetation, water bodies, soil, etc., which usually come from sensors with medium spatial resolution, such as Landsat series satellites. Medium-high resolution forest age remote sensing products are used to estimate the age or growth stage of vegetation in the target area using multi-temporal remote sensing image data through specific algorithms or models, such as inferring by analyzing the changes in vegetation index over time.

[0034] For example, the cloud-free medium-to-high resolution remote sensing image may be Landsat80 LI / TIRS, with a resolution of 30 meters; the medium-to-low resolution vegetation net primary productivity remote sensing product may be MODIS NPP product, with a resolution of 1 kilometer; the medium-to-high resolution land classification remote sensing product may be ESRI land classification product, with a resolution of 10 meters; and the medium-to-high resolution forest age remote sensing product may be a forest age map generated by combining remote sensing images with ground survey data, with a resolution of 30 meters or higher.

[0035] 102. Based on medium- and high-resolution land classification remote sensing products, determine the measured pixel points in the medium- and low-resolution vegetation net primary productivity remote sensing products, and extract the medium- and low-resolution vegetation net primary productivity of each measured pixel point.

[0036] Among them, the quasi-measured pixel points are used to characterize the pixel points with uniform vegetation in the medium- and low-resolution vegetation net primary productivity remote sensing products. Since medium- and high-resolution land classification remote sensing products are used to provide detailed land cover classification information of the target area, including different types of vegetation, water bodies, soil, etc., the vegetation cover type, distribution and area information of the target area can be extracted from the medium- and high-resolution land classification remote sensing products, and the distribution areas of the main vegetation types (such as forests, grasslands, farmlands, etc.) can be identified.

[0037] Since pixels with uniform vegetation can usually represent the typical growth state of a certain type of vegetation under specific environmental conditions and have good representativeness, in the medium and low resolution vegetation net primary productivity remote sensing products, the pixels located in the uniform vegetation distribution area in the medium and high resolution land classification remote sensing products are selected as the quasi-measured pixel points, and for each selected quasi-measured pixel point, the corresponding vegetation net primary productivity is extracted from the medium and low resolution vegetation net primary productivity remote sensing products as the input data for subsequent inversion, which helps to reduce the impact of factors such as mixed vegetation types and differences in growth conditions on the subsequent inversion results. The corresponding pixel points with uniform vegetation can be determined by land cover classification analysis, vegetation coverage analysis or topography analysis. For example, the vegetation coverage analysis method can be used to calculate the vegetation coverage in the target area and select pixel points in blocks with high vegetation coverage and uniform distribution.

[0038] It should be noted that, in this embodiment, for the selection of quasi-measured pixel points, the initial pixel points can be determined by first using the random proportion principle or stratified sampling, and then the initial pixel points with uniform vegetation are selected as the quasi-measured pixel points. This can not only ensure the diversity and representativeness of the quasi-measured pixel points, but also improve the accuracy and stability of subsequent inversion, while also improving the efficiency of selecting quasi-measured pixel points.

[0039] 103. The vegetation index of each pixel is calculated based on the band data corresponding to the cloud-free medium-resolution remote sensing image, and the vegetation index corresponding to the measured pixel of each class is extracted.

[0040] It should be noted that, in this embodiment, the vegetation index is used to characterize an index that can reflect the photosynthesis and carbon sequestration capacity of vegetation, including but not limited to the ratio vegetation index, enhanced vegetation index, soil-adjusted vegetation index, normalized vegetation index, difference vegetation index, optimized soil-adjusted vegetation index, nitrogen reflectance index, vegetation attenuation index, structure-insensitive pigment vegetation index and green normalized vegetation index, etc. It should be noted that the vegetation index can be all of the above indexes or some of the above indexes, which is not limited in this embodiment.

[0041] Since the calculation formulas corresponding to different vegetation indices are constructed based on the band data corresponding to the cloud-free medium and high-resolution remote sensing image, according to the definition of the required vegetation index, the corresponding band (such as near infrared, red light, blue light, etc.) is selected from the cloud-free medium and high-resolution remote sensing image, and the band data of each pixel point is calculated according to its calculation formula to generate a vegetation index map of the entire image, that is, a vegetation index map corresponding to the cloud-free medium and high-resolution remote sensing image, and the vegetation index map contains the vegetation index of each pixel point in the cloud-free medium and high-resolution remote sensing image. Since the subsequent inversion needs to use the medium and low-resolution vegetation net primary productivity and the corresponding vegetation index of the measured pixel point in step 102, the vegetation index value of each measured pixel point can be correspondingly extracted from the generated vegetation index map as another input data for the subsequent inversion.

[0042] 104. The medium- and high-resolution net primary productivity of vegetation at each pixel point in the cloud-free medium- and high-resolution remote sensing image is inverted using the medium- and low-resolution net primary productivity of vegetation at each pixel point in each class and the vegetation index at each pixel point in each class.

[0043] Based on the low- and medium-resolution net primary productivity of vegetation of the class-measured pixel points provided in step 102 and the vegetation index corresponding to the class-measured pixel points provided in step 103, an inversion model can be established, and the inversion model can be a multiple linear regression model, a support vector machine, a neural network, etc. The inversion model is used to estimate the net primary productivity of vegetation at the medium- and high-resolution of the pixel points through the vegetation index, that is, the medium- and high-resolution net primary productivity of vegetation. The inversion model can be specifically trained using the vegetation index corresponding to the class-measured pixel points and the low- and medium-resolution net primary productivity of vegetation, that is, the low- and medium-resolution net primary productivity of vegetation is used as the target variable, the vegetation index corresponding to the class-measured pixel points is used as the feature variable, and the generalization ability of the inversion model is evaluated by cross-validation and the like. The trained inversion model is applied to each pixel point of the cloud-free medium- and high-resolution remote sensing image, and the medium- and high-resolution net primary productivity of vegetation of each pixel point can be calculated according to the vegetation index corresponding to each pixel point.

[0044] 105. Based on medium- and high-resolution forest age remote sensing products, the carbon sink of the medium- and high-resolution vegetation net primary productivity of each pixel in the cloud-free medium- and high-resolution remote sensing images is calculated to obtain the carbon sink amount of each pixel in the cloud-free medium- and high-resolution remote sensing images.

[0045] In actual situations, forest age is one of the important factors affecting the net primary productivity of vegetation. Generally speaking, with the increase of forest age, the biomass, leaf area, photosynthesis efficiency, etc. of vegetation will change significantly, thus affecting its net primary productivity. For example, due to the vigorous growth, the net primary productivity of young forests may be higher; while mature forests and over-mature forests have larger biomass, but due to the decline in growth rate, their net primary productivity may be lower than that of young forests. The medium- and high-resolution forest age remote sensing products are used to provide the age or growth stage of vegetation in the target area. Therefore, the forest age distribution information of the target area can be extracted from the medium- and high-resolution forest age remote sensing products to obtain the forest age corresponding to each pixel point. In order to reduce the impact of forest age on the net primary productivity of vegetation, the carbon sink conversion factor of each pixel point in the cloudless medium- and high-resolution remote sensing image can be determined first. The carbon sink conversion factor is used to characterize the proportion of the net primary productivity of vegetation per unit area of ​​the same vegetation type under different forest age conditions converted into carbon sinks. By using this carbon sink conversion factor to calculate the carbon sink of the net primary productivity of vegetation at each pixel point in the cloud-free medium and high resolution remote sensing image, the carbon sink amount of each pixel point in the cloud-free medium and high resolution remote sensing image can be obtained.

[0046] Specifically, the carbon sink conversion factors corresponding to different vegetation types under different forest age conditions can be constructed in advance based on existing literature, experimental data, model simulation or expert experience. They can be in tabular form (such as fixed coefficients or functional relationships corresponding to different forest age stages) or in model form (such as a multivariate regression model based on factors such as forest age, vegetation type, and climatic conditions). This embodiment does not limit this. By determining the vegetation type in a cloudless medium- and high-resolution remote sensing image, the corresponding carbon sink conversion factor can be determined by the vegetation type. The carbon sink amount is equal to the product of the medium- and high-resolution vegetation net primary productivity and the carbon sink conversion factor. Therefore, by multiplying the carbon sink conversion factor of each pixel point with the carbon sink conversion factor, the carbon sink amount of each pixel point in the cloudless medium- and high-resolution remote sensing image can be obtained, that is, a high-resolution carbon sink distribution map covering the target area.

[0047] Based on the above Figure 1It can be seen from the implementation method that, through the above technical scheme, a remote sensing-based rapid carbon sink monitoring method provided by the present application is that when it is necessary to monitor vegetation carbon sinks, first, multi-source remote sensing data corresponding to the target area is obtained, and the multi-source remote sensing data includes cloud-free medium and high-resolution remote sensing images, medium and low-resolution vegetation net primary productivity remote sensing products, medium and high-resolution land classification remote sensing products and medium and high-resolution forest age remote sensing products. Then, based on the medium and high-resolution land classification remote sensing products, the class-measured pixel points are determined in the medium and low-resolution vegetation net primary productivity remote sensing products, and the medium and low-resolution vegetation net primary productivity of each class-measured pixel point is extracted. The class-measured pixel points are used to characterize the medium and low-resolution vegetation net primary productivity. The pixels with uniform vegetation in the productivity remote sensing products are selected. At the same time, the vegetation index of each pixel is calculated based on the band data corresponding to the cloud-free medium and high-resolution remote sensing images, and the vegetation index corresponding to the measured pixel points of each class is extracted. Then, the medium and low-resolution vegetation net primary productivity of each class measured pixel point and the vegetation index of each class measured pixel point are used to invert the medium and high-resolution vegetation net primary productivity of each pixel point in the cloud-free medium and high-resolution remote sensing images. Finally, based on the medium and high-resolution forest age remote sensing products, the carbon sink of the medium and high-resolution vegetation net primary productivity of each pixel point in the cloud-free medium and high-resolution remote sensing images is calculated to obtain the carbon sink amount of each pixel point in the cloud-free medium and high-resolution remote sensing images. Through the technical solution provided in this application, more detailed vegetation structure and growth status information can be obtained from cloud-free medium and high-resolution remote sensing images, and quasi-measured pixel points that effectively reflect the vegetation carbon sink characteristics of the target area can be selected from medium and low-resolution vegetation net primary productivity remote sensing products and medium and high-resolution land classification remote sensing products. The medium and high-resolution vegetation net primary productivity can be inverted using the vegetation index obtained from the cloud-free medium and high-resolution remote sensing images, and the accurate medium and high-resolution vegetation net primary productivity can be estimated. The carbon sink of the medium and high-resolution vegetation net primary productivity can then be calculated based on the medium and high-resolution forest age remote sensing products. The impact of heterogeneity on carbon sink monitoring is comprehensively considered, and high-precision, rapid, and comprehensive monitoring of vegetation carbon sinks is achieved, thereby effectively improving the efficiency and accuracy of carbon sink monitoring, which is beneficial to the research and management of the carbon cycle.

[0048] Further, the preferred embodiment of the present application is in the above Figure 1 Based on this, a detailed description of the process of rapid carbon sink monitoring based on remote sensing is given. The specific steps are as follows: Figure 2 As shown, including:

[0049] 201. Obtain multi-source remote sensing data corresponding to the target area.

[0050] This step is combined with the description of step 104 in the above method, and the same contents are not repeated here.

[0051] 202. The initial pixel points are selected in the low and medium resolution vegetation net primary productivity remote sensing products according to the random proportion principle.

[0052] In this step, the random proportion principle can be specifically set to a specified random proportion according to the research needs and the total number of pixels of the medium and low resolution vegetation net primary productivity remote sensing products. For example, if there are a total of 10,000 pixels and it is planned to select 1% of the pixels as the initial pixels, 100 initial pixels need to be randomly selected. Use a computer programming language (such as Python, R, MATLAB, etc.) to generate a uniformly distributed random number sequence with the same length as the total number of pixels. The value range of the sequence should correspond to the pixel index range. For example, for 10,000 pixels, the value of the random number sequence should be between 0-9999. The random number sequence is regarded as a pixel index, and the pixel points at the corresponding position are extracted from the medium and low resolution vegetation net primary productivity remote sensing products according to the index value as the initial pixel points.

[0053] Random sampling based on the random proportion principle is used to obtain representative initial pixel points, which can accurately characterize the net primary productivity characteristics of the target area. It helps to ensure that the extracted initial pixel points can reflect the distribution of net primary productivity under the influence of different terrain, climate, soil type and other factors, thereby avoiding sample bias caused by human selection and ensuring the accuracy and comprehensiveness of subsequent inversion.

[0054] 203. The initial pixel points with uniform vegetation in the medium and high resolution land classification remote sensing products are regarded as the quasi-measured pixel points.

[0055] In this step, since the pixel points with uniform vegetation can usually represent the typical growth state of a certain type of vegetation under specific environmental conditions and have good representativeness, after the initial pixel points are determined in the medium and low resolution vegetation net primary productivity remote sensing products, the measured pixel points can be selected from the initial pixel points. Specifically, according to the medium and high resolution land classification remote sensing products, the uniformity of vegetation types in a certain neighborhood (such as 3x3, 5x5 windows) around each initial pixel point is evaluated. For example, the proportion of vegetation types, diversity index, homogeneity index and other indicators are calculated to quantify vegetation uniformity. By setting a threshold corresponding to the quantitative index, when the vegetation uniformity index in the neighborhood exceeds the threshold, the vegetation of the initial pixel point is considered to be uniform, and it can be marked as a measured pixel point. And retain the longitude and latitude of each measured pixel point in the medium and low resolution vegetation net primary productivity remote sensing products.

[0056] 204. Extract the low- and medium-resolution net primary productivity of vegetation of each class from the low- and medium-resolution net primary productivity remote sensing products based on the longitude and latitude of each class's measured pixel point.

[0057] In this step, the latitude and longitude of each class measured pixel point in the medium and low resolution vegetation net primary productivity remote sensing product retained in step 203 can be used to locate the corresponding class measured pixel point in the medium and low resolution vegetation net primary productivity remote sensing product, and extract the medium and low resolution vegetation net primary productivity of the class measured pixel point. The extraction of the medium and low resolution vegetation net primary productivity can be achieved by spatial query or index operation in combination with step 202.

[0058] 205. The vegetation index of each pixel is calculated based on the band data corresponding to the cloud-free medium- and high-resolution remote sensing images, and the vegetation index corresponding to the measured pixel of each class is extracted.

[0059] This step is combined with the description of step 103 in the above method, and the same content will not be repeated here. It should be noted that since different vegetation indices are constructed based on the band data corresponding to the cloud-free medium-resolution remote sensing image, and each vegetation index is a factor reflecting the vegetation growth status and carbon sink capacity, therefore, it is possible to construct a factor system of vegetation growth status and carbon sink capacity based on the cloud-free medium-resolution remote sensing image, that is, select multiple vegetation indices, and use the factor system to reflect vegetation characteristics from multiple angles and levels, such as chlorophyll content, leaf coverage, water status, photosynthetic activity, etc., thereby improving the accuracy and reliability of monitoring dynamic changes in vegetation.

[0060] For example, assuming that the cloud-free medium-high resolution remote sensing image is a Landsdat8 remote sensing image, the calculation formulas for each factor are as follows in Table 1:

[0061]

[0062]

[0063] Table 1

[0064] Among them, C1=6, C2=7.5, L1=1, L2=0.5; B2, B3, B4, B5 represent the blue band, green band, red band and near-infrared band of Landsat8 respectively.

[0065] Since the above calculation is for the vegetation index corresponding to each pixel point in the cloud-free medium-resolution remote sensing image, in order to facilitate the inversion of steps 206-207 based on the quasi-measured pixel points, the vegetation index corresponding to each quasi-measured pixel point can be extracted in combination with the latitude and longitude of the quasi-measured pixel point.

[0066] 206. The low- and medium-resolution net primary productivity of vegetation at each measured pixel point and the vegetation index corresponding to each measured pixel point are used as sample sets, and the inversion model is trained using the multivariate stepwise regression algorithm.

[0067] In this step, the inversion model is used to estimate the medium- and high-resolution net primary productivity of vegetation at the pixel point according to the given vegetation index. Since there is a significant statistical relationship between the vegetation index and the net primary productivity, that is, the more luxuriant the vegetation is and the higher the net primary productivity of vegetation is, the higher the corresponding vegetation index is. In other words, the vegetation index can be used as an indirect, easily accessible and spatially continuous proxy variable for the net primary productivity of vegetation. Therefore, the medium- and low-resolution net primary productivity of vegetation at each class of measured pixel points and the vegetation index corresponding to each class of measured pixel points can be used as a sample set, that is, the medium- and low-resolution net primary productivity of vegetation at each class of measured pixel points is matched with its corresponding vegetation index to form a sample data pair. Each sample data pair includes a medium- and low-resolution net primary productivity of vegetation and a set (possibly including multiple) of vegetation indices. This step preferably uses the multivariate stepwise regression algorithm as the inversion model, which is a method for automatically selecting important independent variables (here, the vegetation index), and gradually adding or removing independent variables through iteration to minimize the model residual sum of squares and control the model complexity.

[0068] Specifically, the medium- and low-resolution net primary productivity of vegetation of each class-measured pixel point and the vegetation index corresponding to each class-measured pixel point are used as sample sets, and the multivariate stepwise regression algorithm is used to train the inversion model, and the inversion model is used to output the medium- and high-resolution net primary productivity of vegetation of the pixel point, including: constructing a sample set based on the target variable and the characteristic variable, the target variable is the medium- and low-resolution net primary productivity of vegetation of the class-measured pixel point, and the characteristic variable is the vegetation index corresponding to the class-measured pixel point; preliminarily constructing the inversion model using the multivariate stepwise regression algorithm and the sample set; dividing the sample set into a training set and a validation set according to a preset ratio, and training and validating the inversion model based on the training set and the validation set, respectively, until the inversion model meets the preset standard.

[0069] Among them, the low- and medium-resolution net primary productivity of vegetation at the measured pixel points is used as the target variable (y), and the corresponding vegetation index is used as the characteristic variable (X), so that each sample contains a target variable value (y) and a set of characteristic variable values ​​(X), and the target variable (NPP) is matched with the characteristic variable (vegetation index) one by one to form a sample set. The preset ratio can be pre-customized according to the total number of sample sets, such as 5:1, 6:2, etc. The sample set is randomly divided into a training set and a validation set according to the preset ratio. The training set is used for model training and parameter optimization, and the validation set is used to evaluate the generalization ability of the model. The training set is used to train the preliminarily constructed multivariate stepwise regression model, and the optimal independent variable combination and model parameters are determined by forward selection, backward elimination or mixed strategies in the stepwise regression process. The trained model is verified using the validation set, and the prediction performance indicators (such as the determination coefficient R 2 , root mean square error RMSE, etc.). Evaluate whether the model meets the preset standards (such as R2 Threshold, RMSE threshold, etc.). If the performance of the model on the validation set does not meet the preset standards, you can adjust the model parameters (such as regularization parameters), change the independent variable screening strategy, or try other regression algorithms, and retrain and verify the model until the performance of the model on the validation set meets the preset standards. Based on the model that meets the preset standards, determine the final independent variable combination and model parameters. At this time, save the model in an appropriate format (such as .pk l, .job lib, .h5, etc.) as an inversion model for subsequent applications.

[0070] 207. The vegetation index corresponding to each pixel is input into the inversion model to obtain the medium- and high-resolution net primary productivity of vegetation at each pixel in the cloud-free medium- and high-resolution remote sensing image.

[0071] Since the inversion model has been trained in step 206, in this step, for each pixel point in the cloud-free medium-high resolution remote sensing image, the vegetation index corresponding to each pixel point can be input into the inversion model, so that it outputs the medium-high resolution vegetation net primary productivity of each pixel point in the cloud-free medium-high resolution remote sensing image. Specifically, a medium-high resolution vegetation net primary productivity distribution map can be generated for intuitive display and further analysis.

[0072] 208. Determine the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium- and high-resolution remote sensing image based on the medium- and high-resolution forest age remote sensing products.

[0073] Among them, the carbon sink conversion factor is used to characterize the ratio of the net primary productivity per unit area of ​​the same vegetation type converted into carbon sink under different forest age conditions. For the same vegetation type, the carbon sink efficiency multiplied by the medium and high resolution net primary productivity of vegetation is equal to the carbon sink. Therefore, the carbon sink conversion factor can be calculated based on the carbon sink efficiency. For example, the basic carbon sink efficiency of different vegetation types can be determined in advance based on existing literature, experimental data, model simulation or expert experience. The basic carbon sink efficiency refers to the ratio of the net primary productivity per unit area of ​​the same vegetation type converted into carbon sink under ideal or typical growth conditions. By introducing the forest age, that is, by the change of the basic carbon sink efficiency at different ages or growth stages, it is corrected to obtain the carbon sink conversion factor. Specifically, the basic carbon sink efficiency of each pixel point in the cloudless medium and high resolution remote sensing image is determined according to the vegetation type corresponding to each pixel point in the cloudless medium and high resolution remote sensing image; the basic carbon sink efficiency of each pixel point in the cloudless medium and high resolution remote sensing image is corrected according to the forest age of each pixel point in the medium and high resolution forest age remote sensing products, and the carbon sink conversion factor corresponding to each pixel point in the cloudless medium and high resolution remote sensing image is obtained.

[0074] In this step, a carbon sink efficiency correction model for different vegetation types at different forest age stages can be constructed based on ecological principles, measured data, model simulation or expert knowledge. The carbon sink efficiency correction model can be in the form of a table (such as correction coefficients or functional relationships corresponding to different forest age stages), a curve (such as a power function, exponential function, S-shaped curve, etc.) or a complex mathematical model (such as a growth model, a carbon cycle model, etc.). The basic carbon sink efficiency of each pixel point in the cloud-free high-resolution remote sensing image is combined with its corresponding forest age information and corrected through the constructed forest age-carbon sink efficiency correction model to obtain the corrected carbon sink efficiency of each pixel point under the current forest age condition.

[0075] 209. Based on the carbon sink conversion factor corresponding to each pixel in the cloud-free medium- and high-resolution remote sensing image, the carbon sink of the net primary productivity of vegetation at each pixel in the cloud-free medium- and high-resolution remote sensing image is calculated to obtain the carbon sink amount of each pixel in the cloud-free medium- and high-resolution remote sensing image.

[0076] In this step, the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium-high resolution remote sensing image obtained in step 208 can be multiplied by the net primary productivity of vegetation at each pixel point in the cloud-free medium-high resolution remote sensing image, thereby obtaining the carbon sink amount of each pixel point in the cloud-free medium-high resolution remote sensing image. Specifically, a high-resolution carbon sink distribution map covering the target area can be generated.

[0077] Furthermore, since the present embodiment monitors carbon sinks based on remote sensing images, in order to verify the credibility of the carbon sinks monitored by the present embodiment, in order to avoid the large error of the carbon sinks affecting the subsequent research and management of the carbon cycle. Specifically, it includes: obtaining similar areas of the target area; if the actual measurement condition is that a flux tower is deployed, then extracting the carbon flux measured by the flux tower; if the actual measurement condition is that the box method can be implemented, then extracting the carbon flux measured by the box method; using the preset verification rules to verify the carbon sink of each pixel point in the cloud-free medium and high-resolution remote sensing image based on the carbon flux measured by the flux tower or the carbon flux measured by the box method, and obtaining the verification results, which are used to characterize the credibility of the carbon sink of each pixel point in the cloud-free medium and high-resolution remote sensing image.

[0078] Among them, similar areas are areas with the same climate zone, similar vegetation types and measured conditions as the target area, and the measured conditions include at least the deployment of flux towers and the implementation of the box method. Because in the same climate zone, the vegetation growth conditions faced by similar vegetation types are roughly the same, and the measured conditions can reflect the actual carbon sink. Therefore, similar areas of the target area can be deployed in advance for verification based on measured data. Specifically, the first round of candidate areas with similar climate zoning can be found according to the climate type of the target area (such as humid tropical, temperate monsoon, arid areas, etc.), and then the second round of candidate areas with similar vegetation types are determined from the initial candidate areas through remote sensing data, vegetation classification maps, plant community surveys, etc., and whether the second round of candidate areas have the same field observation facilities as the target area, such as the deployment of flux towers or conditions suitable for measuring carbon fluxes by the box method.

[0079] After determining similar areas, if flux towers are deployed in similar areas, carbon flux can be obtained through public data sets, cooperation with research institutions, field surveys, etc. If the box method can be implemented in similar areas, the aboveground and underground biomass measured by the box method is obtained and the carbon flux is estimated. The measured carbon flux is spatially matched with the carbon sink of each pixel point in the cloud-free medium-resolution remote sensing image to ensure the correspondence between the pixel point and the measured point in time and space.

[0080] The pre-verification rule is used to characterize the comparison rule based on the difference between at least one of the correlation coefficient, root mean square error and deviation percentage corresponding to the carbon flux and the carbon sink and the set difference threshold. When the difference is less than or equal to the set difference threshold, it means that the credibility of the carbon sink is high, otherwise it means that the credibility is low.

[0081] Furthermore, as a response to the above Figure 1-2 The implementation of the method embodiment shown in the figure, the embodiment of the present application provides a carbon sink rapid monitoring device based on remote sensing, which is used to improve the efficiency and accuracy of carbon sink monitoring. The embodiment of the device corresponds to the aforementioned method embodiment. For ease of reading, this embodiment will no longer repeat the details of the aforementioned method embodiment one by one, but it should be clear that the device in this embodiment can correspond to all the contents of the aforementioned method embodiment. Specifically, Figure 3 As shown, the device comprises:

[0082] The first acquisition unit 31 is used to acquire multi-source remote sensing data corresponding to the target area, wherein the multi-source remote sensing data includes cloud-free medium-high resolution remote sensing images, medium-low resolution vegetation net primary productivity remote sensing products, medium-high resolution land classification remote sensing products, and medium-high resolution forest age remote sensing products;

[0083] A determination unit 32 is used to determine the quasi-measured pixel points in the medium- and low-resolution vegetation net primary productivity remote sensing product based on the medium- and high-resolution land classification remote sensing product of the first acquisition unit 31, and extract the medium- and low-resolution vegetation net primary productivity of each of the quasi-measured pixel points, wherein the quasi-measured pixel points are used to characterize the pixel points with uniform vegetation in the medium- and low-resolution vegetation net primary productivity remote sensing product;

[0084] A calculation unit 33 is used to calculate the vegetation index of each pixel point based on the band data corresponding to the cloud-free high-resolution remote sensing image of the first acquisition unit 31, and extract the vegetation index corresponding to each of the quasi-measured pixel points;

[0085] The inversion unit 34 is used to invert the medium- and high-resolution net primary productivity of vegetation of each of the quasi-measured pixel points of the determination unit 32 and the vegetation index of each of the quasi-measured pixel points of the calculation unit 33 to obtain the medium- and high-resolution net primary productivity of vegetation of each pixel point in the cloud-free medium- and high-resolution remote sensing image;

[0086] The processing unit 35 is used to calculate the carbon sink of the medium-high resolution vegetation net primary productivity of each pixel point in the cloud-free medium-high resolution remote sensing image of the inversion unit 34 based on the medium-high resolution forest age remote sensing product, so as to obtain the carbon sink amount of each pixel point in the cloud-free medium-high resolution remote sensing image.

[0087] Further, such as Figure 4 As shown, the determining unit 32 includes:

[0088] A selection module 321 is used to select initial pixel points from the medium and low resolution vegetation net primary productivity remote sensing products according to a random proportion principle;

[0089] A processing module 322 is used to use the initial pixel points of the selection module 321 with uniform vegetation in the medium- and high-resolution land classification remote sensing products as quasi-measured pixel points;

[0090] The extraction module 323 is used to extract the medium and low resolution net primary productivity of vegetation of each of the type measured pixel points in the medium and low resolution net primary productivity of vegetation remote sensing products according to the longitude and latitude of each of the type measured pixel points in the processing module 322.

[0091] Further, such as Figure 4 As shown, the inversion unit 34 comprises:

[0092] The training module 341 is used to use the medium- and low-resolution net primary productivity of vegetation of each of the measured pixel points and the vegetation index corresponding to each of the measured pixel points as a sample set, and train an inversion model using a multivariate stepwise regression algorithm, wherein the inversion model is used to estimate the medium- and high-resolution net primary productivity of vegetation of the pixel point according to a given vegetation index;

[0093] The processing module 342 is used to input the vegetation index corresponding to each pixel point into the inversion model to obtain the medium- and high-resolution net primary productivity of vegetation at each pixel point in the cloud-free medium- and high-resolution remote sensing image.

[0094] Further, such as Figure 4 As shown, the training module 341 is specifically used for:

[0095] The sample set is constructed based on a target variable and a characteristic variable, wherein the target variable is the low- and medium-resolution net primary productivity of vegetation of the class measured pixel points, and the characteristic variable is the vegetation index corresponding to the class measured pixel points;

[0096] Preliminarily constructing the inversion model using the multivariate stepwise regression algorithm and the sample set;

[0097] The sample set is divided into a training set and a verification set according to a preset ratio, and the inversion model is trained and verified based on the training set and the verification set respectively until the inversion model meets the preset standard.

[0098] Further, such as Figure 4 As shown, the processing unit 35 includes:

[0099] A determination module 351 is used to determine the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium-high resolution remote sensing image according to the medium-high resolution forest age remote sensing product, wherein the carbon sink conversion factor is used to characterize the ratio of the net primary productivity per unit area of ​​vegetation of the same vegetation type under different forest age conditions converted into carbon sink amount;

[0100] The correction module 352 is used to correct the net primary productivity of vegetation at each pixel point in the cloud-free medium-resolution remote sensing image based on the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium-resolution remote sensing image of the determination module 351, so as to obtain the carbon sink amount of each pixel point in the cloud-free medium-resolution remote sensing image.

[0101] Further, such as Figure 4 As shown, the determination module 351 is specifically used to:

[0102] Determining the basic carbon sink efficiency of each pixel point in the cloud-free medium-high resolution remote sensing image according to the vegetation type corresponding to each pixel point in the cloud-free medium-high resolution remote sensing image;

[0103] The basic carbon sink efficiency of each pixel point in the cloud-free medium and high resolution remote sensing image is corrected according to the forest age of each pixel point in the medium and high resolution forest age remote sensing product to obtain the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium and high resolution remote sensing image.

[0104] Further, such as Figure 4 As shown, the device also includes:

[0105] A second acquisition unit 36 ​​is used to acquire a similar area to the target area, wherein the similar area is an area with the same climate zone and similar vegetation type as the target area and has actual measurement conditions, wherein the actual measurement conditions include at least one of the deployment of a flux tower and the implementation of a box method;

[0106] an extraction unit 37, configured to extract the carbon flux measured by the flux tower if the actual measurement condition of the second acquisition unit 36 ​​is that a flux tower is deployed;

[0107] The extraction unit 37 is further used to extract the carbon flux measured by the box method if the actual measurement condition of the second acquisition unit 36 ​​is that the box method can be implemented;

[0108] The verification unit 38 is used to verify the carbon sequestration amount of each pixel point in the cloud-free high-resolution remote sensing image based on the carbon flux measured by the flux tower of the extraction unit 37 or the carbon flux measured by the box method using preset verification rules to obtain a verification result, and the verification result is used to characterize the credibility of the carbon sequestration amount of each pixel point in the cloud-free high-resolution remote sensing image.

[0109] Furthermore, an embodiment of the present application further provides a storage medium, wherein the storage medium is used to store a computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the above Figure 1-2 The rapid carbon sink monitoring method based on remote sensing described in.

[0110] Furthermore, the embodiment of the present application also provides a processor, the processor is used to run a program, wherein the program executes the above Figure 1-2 The rapid carbon sink monitoring method based on remote sensing described in.

[0111] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0112] It is understandable that the related features in the above methods and devices can be referenced to each other. In addition, the "first", "second" and the like in the above embodiments are used to distinguish the embodiments, but do not represent the advantages and disadvantages of the embodiments.

[0113] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0114] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the application is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the application described herein, and the description of the specific language above is for the purpose of disclosing the best mode of implementation of the application.

[0115] In addition, the memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0116] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0117] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0118] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0119] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0120] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0121] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0122] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0123] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0124] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0125] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A carbon sink rapid monitoring method based on remote sensing, characterized in that: The method comprises: Acquire multi-source remote sensing data corresponding to the target area, wherein the multi-source remote sensing data includes cloud-free medium-high resolution remote sensing images, medium-low resolution vegetation net primary productivity remote sensing products, medium-high resolution land classification remote sensing products, and medium-high resolution forest age remote sensing products; Based on the medium-to-high resolution land classification remote sensing product, the quasi-measured pixel points are determined in the medium-to-low resolution vegetation net primary productivity remote sensing product, and the medium-to-low resolution vegetation net primary productivity of each quasi-measured pixel point is extracted. The quasi-measured pixel point is used to characterize the pixel point with uniform vegetation in the medium-to-low resolution vegetation net primary productivity remote sensing product. Specifically, the initial pixel point is selected in the medium-to-low resolution vegetation net primary productivity remote sensing product according to the random proportion principle; the initial pixel point with uniform vegetation in the medium-to-high resolution land classification remote sensing product is used as the quasi-measured pixel point; the medium-to-low resolution vegetation net primary productivity of each quasi-measured pixel point in the medium-to-low resolution vegetation net primary productivity remote sensing product is extracted according to the longitude and latitude of each quasi-measured pixel point; Calculating the vegetation index of each pixel point based on the band data corresponding to the cloud-free high-resolution remote sensing image, and extracting the vegetation index corresponding to each of the measured pixel points; The medium- and high-resolution net primary productivity of vegetation of each of the class-measured pixel points and the vegetation index of each of the class-measured pixel points are used for inversion to obtain the medium- and high-resolution net primary productivity of vegetation of each pixel point in the cloud-free medium- and high-resolution remote sensing image. Specifically, the medium- and low-resolution net primary productivity of vegetation of each of the class-measured pixel points and the vegetation index corresponding to each of the class-measured pixel points are used as a sample set, and an inversion model is trained using a multivariate stepwise regression algorithm. The inversion model is used to estimate the medium- and high-resolution net primary productivity of vegetation of the pixel point according to a given vegetation index; the vegetation index corresponding to each of the pixel points is input into the inversion model to obtain the medium- and high-resolution net primary productivity of vegetation of each pixel point in the cloud-free medium- and high-resolution remote sensing image; The carbon sink calculation is performed on the medium-high resolution vegetation net primary productivity of each pixel point in the cloudless medium-high resolution remote sensing image based on the medium-high resolution forest age remote sensing product to obtain the carbon sink amount of each pixel point in the cloudless medium-high resolution remote sensing image. Specifically, the carbon sink conversion factor corresponding to each pixel point in the cloudless medium-high resolution remote sensing image is determined according to the medium-high resolution forest age remote sensing product, and the carbon sink conversion factor is used to characterize the proportion of the net primary productivity of vegetation per unit area of ​​the same vegetation type under different forest age conditions converted into carbon sink amount; based on the carbon sink conversion factor corresponding to each pixel point in the cloudless medium-high resolution remote sensing image, the vegetation net primary productivity of each pixel point in the cloudless medium-high resolution remote sensing image is corrected to obtain the carbon sink amount of each pixel point in the cloudless medium-high resolution remote sensing image.

2. The method according to claim 1, characterized in that The medium and low resolution net primary productivity of vegetation of each of the measured pixel points and the vegetation index corresponding to each of the measured pixel points are used as a sample set, and the inversion model is trained using a multivariate stepwise regression algorithm. The inversion model is used to output the medium and high resolution net primary productivity of vegetation of the pixel points, including: The sample set is constructed based on a target variable and a characteristic variable, wherein the target variable is the low- and medium-resolution net primary productivity of vegetation of the class measured pixel points, and the characteristic variable is the vegetation index corresponding to the class measured pixel points; Preliminarily constructing the inversion model using the multivariate stepwise regression algorithm and the sample set; The sample set is divided into a training set and a verification set according to a preset ratio, and the inversion model is trained and verified based on the training set and the verification set respectively until the inversion model meets the preset standard.

3. The method according to claim 1, characterized in that Determining the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium- and high-resolution remote sensing image according to the medium- and high-resolution forest age remote sensing product includes: Determining the basic carbon sink efficiency of each pixel point in the cloud-free medium-high resolution remote sensing image according to the vegetation type corresponding to each pixel point in the cloud-free medium-high resolution remote sensing image; The basic carbon sink efficiency of each pixel point in the cloud-free medium and high resolution remote sensing image is corrected according to the forest age of each pixel point in the medium and high resolution forest age remote sensing product to obtain the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium and high resolution remote sensing image.

4. The method according to any one of claims 1 to 3, characterized in that The method further comprises: Acquire a similar area to the target area, wherein the similar area is an area having the same climate zone and similar vegetation type as the target area and having actual measurement conditions, wherein the actual measurement conditions at least include any one of deploying a flux tower and being able to implement a box method; If the measured condition is that a flux tower is deployed, then extracting the carbon flux measured by the flux tower; If the measured conditions are such that the box method can be implemented, then the carbon flux measured by the box method is extracted; The carbon sequestration amount of each pixel point in the cloud-free high-resolution remote sensing image is verified based on the carbon flux measured by the flux tower or the carbon flux measured by the box method using preset verification rules to obtain a verification result, which is used to characterize the credibility of the carbon sequestration amount of each pixel point in the cloud-free high-resolution remote sensing image.

5. A carbon sink rapid monitoring device based on remote sensing, characterized in that: The device comprises: A first acquisition unit is used to acquire multi-source remote sensing data corresponding to the target area, wherein the multi-source remote sensing data includes cloud-free medium-high resolution remote sensing images, medium-low resolution vegetation net primary productivity remote sensing products, medium-high resolution land classification remote sensing products, and medium-high resolution forest age remote sensing products; A determination unit is used to determine the quasi-measured pixel points in the medium- and low-resolution vegetation net primary productivity remote sensing product based on the medium- and high-resolution land classification remote sensing product of the first acquisition unit, and extract the medium- and low-resolution vegetation net primary productivity of each of the quasi-measured pixel points, wherein the quasi-measured pixel points are used to characterize the pixel points with uniform vegetation in the medium- and low-resolution vegetation net primary productivity remote sensing product. Specifically, the determination unit includes: a selection module, which is used to select initial pixel points in the medium- and low-resolution vegetation net primary productivity remote sensing product according to a random proportion principle; a processing module, which is used to use the initial pixel points of the selection module with uniform vegetation in the medium- and high-resolution land classification remote sensing product as quasi-measured pixel points; an extraction module, which is used to extract the medium- and low-resolution vegetation net primary productivity of each quasi-measured pixel point in the medium- and low-resolution vegetation net primary productivity remote sensing product according to the latitude and longitude of each quasi-measured pixel point in the processing module; A calculation unit, used to calculate the vegetation index of each pixel point based on the band data corresponding to the cloud-free high-resolution remote sensing image of the first acquisition unit, and extract the vegetation index corresponding to each of the quasi-measured pixel points; An inversion unit is used to perform inversion using the low- and medium-resolution net primary productivity of vegetation of each of the class-measured pixel points of the determination unit and the vegetation index of each of the class-measured pixel points of the calculation unit to obtain the high- and medium-resolution net primary productivity of vegetation of each pixel point in the cloud-free medium- and high-resolution remote sensing image. Specifically, the inversion unit includes: a training module, used to use the low- and medium-resolution net primary productivity of vegetation of each of the class-measured pixel points and the vegetation index corresponding to each of the class-measured pixel points as a sample set, and to train an inversion model using a multivariate stepwise regression algorithm, wherein the inversion model is used to estimate the high- and medium-resolution net primary productivity of vegetation of the pixel point according to a given vegetation index; a processing module, used to input the vegetation index corresponding to each of the pixel points into the inversion model to obtain the high- and medium-resolution net primary productivity of vegetation of each pixel point in the cloud-free medium- and high-resolution remote sensing image; A processing unit is used to perform carbon sink calculation on the medium-high resolution vegetation net primary productivity of each pixel point in the cloud-free medium-high resolution remote sensing image of the inversion unit based on the medium-high resolution forest age remote sensing product, so as to obtain the carbon sink amount of each pixel point in the cloud-free medium-high resolution remote sensing image. Specifically, the processing unit includes: a determination module, used to determine the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium-high resolution remote sensing image based on the medium-high resolution forest age remote sensing product, wherein the carbon sink conversion factor is used to characterize the ratio of the net primary productivity per unit area of ​​vegetation of the same vegetation type under different forest age conditions converted into the carbon sink amount; a correction module, used to correct the vegetation net primary productivity of each pixel point in the cloud-free medium-high resolution remote sensing image based on the carbon sink conversion factor corresponding to each pixel point in the cloud-free medium-high resolution remote sensing image of the determination module, so as to obtain the carbon sink amount of each pixel point in the cloud-free medium-high resolution remote sensing image.

6. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the remote sensing-based carbon sink rapid monitoring method according to any one of claims 1 to 4.

7. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the remote sensing-based carbon sink rapid monitoring method as described in any one of claims 1 to 4.

Images