85 results about "Vegetation type" patented technology

The invention discloses a forest fire risk prediction method based on multi-source data, and belongs to the technical field of forest fire prevention. Aiming at the problem of low prediction precision caused by one-sided information of a single data source and insufficient multi-source data fusion in the prior art, the method is realized by the following steps: acquiring micrometeorological data including temperature and humidity, wind speed and air pressure, and image data including an infrared image and a visible light image; the data of the mountain fire-prone area comprises historical fire frequency, vegetation type and topographic information; uTC + 8 time synchronization and WGS84 coordinate system space calibration are carried out on the data, and missing values and abnormal values are processed; carrying out feature layer fusion by adopting an attention mechanism, and extracting core features such as a temperature and humidity coupling index and vegetation dryness; spatial correlation features are captured through CNN, a time sequence trend is captured through LSTM, a mountain fire occurrence probability is output by using a Sigmoid function after decision-making layer fusion, and a result is calibrated in combination with sub-region features. Through multi-source data deep fusion and spatial-temporal feature collaborative analysis, the accuracy and timeliness of forest fire risk prediction are improved, a new data source can be expanded and accessed, and the method is suitable for a complex forest fire prevention scene.

The invention discloses a road construction carbon emission monitoring method and system, and relates to the technical field of carbon emission monitoring. The method comprises the steps of obtaining earth surface feature information, vegetation types and distribution areas of a construction area, obtaining disturbance areas of different vegetation types, calculating disturbance coefficients, loss coefficients and carbon sink loss amounts of the disturbance areas, calculating the total carbon emission amount of a construction period, constructing an emission evaluation model, and outputting a predicted emission amount of a subsequent construction period in combination with a project scheme. Judging whether the predicted emission exceeds a preset emission threshold value or not, if not, keeping the project scheme unchanged, and if yes, suggesting to modify the project scheme; the carbon sink loss amount caused by vegetation damage can be accurately calculated, so that the total carbon emission amount is more comprehensively and accurately evaluated, and a scientific basis is provided for construction scheme optimization.

The invention discloses a forest fire danger early warning system and method based on meteorological and vegetation data analysis, and relates to the technical field of fire risk prediction and control. According to the forest fire danger early warning system and method based on meteorological and vegetation data analysis, fine grid division is carried out on a target forest area through a high-resolution DEM and a multispectral or hyperspectral image, and vegetation type identification is carried out on each monitoring subunit; calculating a wet-dry cycle index WDI, a monsoon effect index MFI and a dry thunderstorm risk factor DRF by combining hourly wind speed, wind direction, relative humidity, precipitation and vapor pressure difference meteorological data; through extracting canopy height, canopy density, under-forest shrub and grass layer thickness and high VOC tree species coverage rate, fuel continuity coefficient FCI and volatile matter flammability index VEI are calculated, and accurate quantification of continuity and flammability of vertical ladder fuel and horizontal canopy is realized. And a comprehensive fire danger index FFI is constructed and subjected to hierarchical management after judgment, so that the dynamism, the hierarchical property and the pertinence of fire danger early warning are realized, and the forest fire danger prevention and control precision and the patrol resource utilization efficiency are improved.

The present application relates to the technical field of ecological restoration scene modeling, in particular to an ecological restoration scene modeling method and system platform applied to land space planning, comprising the following steps: based on successful ecological restoration technical scheme cases and land units to be restored.The present application realizes modeling in data source selection granularity and expression dimension by normalizing the multi-dimensional attributes between successful ecological restoration technical scheme cases and land units to be restored, constructing source-target land feature fingerprint sets, and combining annual mean temperature, precipitation, soil type, pH value, altitude, native vegetation type and labor cost and other specific attributes.Further, by setting importance coefficients according to the influence degree of each dimension in historical restoration success cases, the present application guides feature difference weight calculation, and obtains technical migration adaptation score by combining normalized difference value calculation, thereby establishing a quantifiable source land adaptation capability evaluation index, and breaking the original limitations of relying on experience judgment.

The application provides a vegetation carbon storage prediction method and device and a storage medium, wherein the method comprises the following steps: extracting a feature set of a to-be-measured region according to a remote sensing image, an elevation model and historical vegetation index information of the to-be-measured region; inputting the feature set into a random forest vegetation classification model for prediction to obtain at least one vegetation type of the to-be-measured region and regional information of each vegetation type in the remote sensing image; performing segmentation processing on the remote sensing image of the to-be-measured region based on a pre-trained vegetation segmentation model to obtain at least one vegetation proportion information of the to-be-measured region and corresponding regional information of each vegetation proportion information in the remote sensing image; and determining the carbon storage of the to-be-measured region according to the vegetation type of the to-be-measured region, the regional information of each vegetation type in the remote sensing image, the vegetation proportion information and the corresponding regional information of each vegetation proportion information in the remote sensing image. The application effectively improves the accuracy of forest carbon storage estimation.

This invention belongs to the field of ecological environment monitoring, specifically a dynamic monitoring method for carbon flux in wetland ecosystems that integrates eddy covariance (ECD) and satellite remote sensing. The method includes the following steps: S1: Monitoring station establishment: Based on the ECD method, flux monitoring stations are established in typical areas of lake wetlands. The selected typical areas should represent the main ecological characteristics and vegetation types of the wetlands to ensure the representativeness and reliability of the monitoring data. This invention combines ECD observation with satellite remote sensing technology to achieve real-time visual monitoring of carbon flux. It provides both data accumulation and visualization, improving the accuracy and reliability of wetland carbon source / sink assessment results. Simultaneously, the visualized spatial distribution of carbon sources / sinks facilitates the detection of changes in carbon sources / sinks in wetland areas, enabling timely detection and response to changes in wetland ecosystems, and scientifically increasing wetland carbon sinks through artificial regulation.

The present application relates to a kind of grassland vegetation coverage measuring device and measuring method, device includes connecting column, lifting mechanism, host computer and camera, host computer is connected with lifting mechanism and camera communication;Connecting column includes vertical connection column and cross bar, column one end is fixed on ground, host computer is installed on column;Lifting mechanism includes lifting block and telescopic machine, the output shaft of telescopic machine connects lifting block, the top of lifting block is equipped with guide shaft, the one end of cross bar is equipped with with guide shaft cooperation installation guide hole, camera is connected with lifting block;Camera changes observation height according to vegetation type.Compared with prior art, the height of camera is controlled by lifting mechanism in the present application, when using, camera hangs under cross bar to collect information, when height needs to be adjusted, telescopic machine can be started by host computer or remote signal control, drive camera to rise or fall, so as to change the effect of camera height control collection range, and the measuring effect is better.

The invention discloses a farmland tree remote sensing image optimization method, relates to the technical field of remote sensing image processing, and solves the problems that in an existing optical image processing technology, a data processing flow is too complex in an optical image processing process, and wrong division and missing division are likely to occur, so that farmland tree images in optical images are not clear, and the image quality is poor. Therefore, the application of optical images in agricultural operation activities is restricted. A farmland tree remote sensing image optimization method comprises a data acquisition stage, a vegetation and non-vegetation distinguishing stage, a vegetation type distinguishing stage, a farmland tree and non-farmland tree distinguishing stage and a farmland tree remote sensing image optimization stage. The method is suitable for the fields of farmland tree remote sensing monitoring, ecology, remote sensing technology, geographic information systems and the like.

The invention discloses a resident-free island vegetation classification method. The method comprises the steps of S1, data acquisition and preprocessing; s2, constructing a random forest model; s3, feature importance evaluation; s4, evaluating feature correlation; s5, feature optimization; s6, precision evaluation; according to the method, the problem that high-precision classification and time sequence dynamic monitoring of non-resident island coastal blue carbon resources in a complex environment are difficult to realize is solved, a remote sensing fine monitoring method is provided, and only satellite remote sensing and ground monitoring data are used; and fine classification and time sequence change detection of resident-free island vegetation types are realized.

A cascade vegetation surface moisture content detection method based on a visual image is characterized by comprising the following steps: step 1, constructing a cascade vegetation surface moisture content detection system based on the visual image; 2, the image acquisition module acquires a vegetation image of a target area; 3, the surface state recognition module quickly recognizes attachments on the surface of the vegetation in the vegetation image, and classifies the vegetation surface state into a saturated water-containing state and a surface dry state; step 4, the vegetation type identification module carries out vegetation type identification on the vegetation image I in the surface dry state, and identifies the biological type of each vegetation in the vegetation image I; and 5, according to the biological category of each vegetation, the moisture content prediction module calls the exclusive moisture content regression prediction model of the corresponding category from the model library to predict the moisture content of the vegetation of the biological category, and outputs a moisture content prediction result of each vegetation.

The invention discloses a data-driven actual evapotranspiration estimation method fusing vegetation biophysical characteristics, which comprises the following steps: firstly, collecting actually measured meteorological and remote sensing data of a flux tower station, and carrying out quality control and energy balance correction to obtain an actual evapotranspiration reference value; then, a PMLV2 model is constructed station by station based on the processed data, and vegetation biophysical characteristic variables such as stomatal conductance are obtained through simulation; summarizing the features with meteorological and remote sensing data to form an input factor set, and summarizing training sets and test sets of all stations according to vegetation types; and finally, taking the factor set as input, constructing an LSTM actual evapotranspiration estimation model by using the training set, and evaluating the precision of the model through the test set. According to the method, vegetation biophysical characteristic variables simulated by the PMLV2 model are fused, so that the adaptability and generalization ability of the data driving model when environmental conditions change are enhanced, and the problems of overfitting and performance reduction when samples are scarce or representativeness is insufficient in a traditional data driving method are effectively relieved.

The utility model discloses a reservoir area hydro-fluctuation belt ecological slope protection structure, and relates to the field of ecological slope protection structures, the reservoir area hydro-fluctuation belt ecological slope protection structure comprises a soil base and a concrete connecting seat, underwater gabion protection feet are attached to the bottom of the side of the soil base at equal intervals, and a gravel permeation area is laid on the outer surface of the soil base; and a vegetation type Reynolds protection pad is fixedly arranged on the outer surface of the gravel permeation area. According to the ecological slope protection structure for the reservoir area hydro-fluctuation belt, through the arrangement of the vegetation type Reynolds protection pad, when the vegetation type Reynolds protection pad covers the outer surface of the gravel permeation area, it can be effectively ensured that the gravel permeation area does not slip off when being impacted by water flow; according to the ecological slope protection structure, the plant seeds can have an excellent soil fixation function during germination and growth, so that the ecological slope protection structure is ensured not to have a large-area soil loss phenomenon during a water swelling period, and the ecological slope protection structure can be ensured to be normally used after being used for a long time.

The present invention relates to a modular grow house system designed for optimizing the cultivation of various vegetation types through advanced environmental control. The system comprises multiple grow chambers, each equipped with sensors to monitor temperature, humidity, light, and soil moisture, and actuators to regulate these conditions. A central control unit, including receiving, monitoring, processing, and regulating units, manages the environmental conditions based on plant profiles stored in digital formats like JSON, XML, or CSV. The modular design features interchangeable panels, standardized sensor and actuator mounts, and quick-connect fittings, allowing flexible configuration and expansion. Safety features, such as temperature and humidity alarms, and a user interface for manual control, enhance the system's functionality. Additionally, the control unit supports remote updates via a cloud platform, ensuring up-to-date plant profiles and control algorithms. This invention provides a scalable, automated solution for controlled environment agriculture, improving productivity and sustainability in various settings.

The application provides a kind of geological disaster post-vegetation recovery monitoring method based on high-resolution remote sensing image, comprising: constructing fusion type "spectrum-vegetation index-texture" feature set;In sunny weather conditions, more evenly select n typical vegetation samples in the target area;Construct bagging type ensemble learning model;Analysis of the obvious differences of arbor, shrub and herbaceous vegetation in vegetation height, single plant vegetation horizontal projection coverage area, root depth and root extension range;By analyzing the vegetation change rate index and vegetation type change rate index of several continuous time phase remote sensing images according to the time phase change information, the vegetation recovery of the target area is quantitatively analyzed, so as to effectively improve the accuracy of pixel classification method, greatly improve the accuracy of quantitative analysis of vegetation in the target area, and greatly reduce the cost of long-term accurate monitoring.

The invention discloses a system and a method for monitoring longitudinal discrete coefficients of river pollutants under vegetation influence. A data acquisition unit acquires parameter data of a vegetation river in real time; the data processing unit performs data preprocessing on the parameter data; the modeling correction unit introduces a vegetation interference factor on the basis of the hydraulic model, constructs a longitudinal discrete coefficient dynamic calculation model, performs adaptive parameter correction on the longitudinal discrete coefficient dynamic calculation model, and outputs a longitudinal discrete coefficient and a pollutant concentration change trend; and the intelligent early warning unit sends out early warning and generates a corresponding treatment scheme based on a preset multi-level threshold triggering mechanism. Vegetation interference factors are introduced and combined with vegetation density, height and resistance coefficients, so that the pollutant discrete monitoring precision is improved; in combination with real-time data acquisition and intelligent early warning, response measures are made for different pollution risks; model coefficients are adjusted according to river vegetation types and hydrological conditions, and the method is suitable for various vegetation-containing water bodies and has wide application prospects in the field of water environment protection and treatment.

The application discloses a kind of estimation method and system of vegetation canopy combustible moisture content based on meteorology and remote sensing data, comprising the following steps: first, combustible moisture content and various meteorological data and remote sensing data and other several kinds of subsidiary data are selected as combustible moisture content estimation data.Meteorological data includes air temperature, relative humidity, rainfall and wind speed.Remote sensing data includes two vegetation indexes: enhanced vegetation index and normalized vegetation index.Subsidiary data includes: root zone soil moisture, vapor pressure difference, drought index, fire weather factor.Then the long time sequence characteristics of meteorological data are extracted.The size of time window is determined first, and the experimental results show that the correlation coefficient of most sites is relatively high under the time window of 90-210 days.90 days, 150 days and 210 days of time window are selected respectively to extract the time characteristics of four kinds of meteorological data.Secondly, the samples of experimental area are divided into five vegetation classifications, which are closed shrub, sparse shrub, multi-tree tropical grassland, tropical savanna and grassland.Finally, the data sets of the five different vegetation types are sequentially adjusted to obtain the respective estimation model.

The invention provides an intelligent remote sensing surveying and mapping method which is characterized by comprising the following steps: acquiring multi-source remote sensing data of a forest region to be measured; performing targeted preprocessing on the multi-source remote sensing data to obtain purified earth surface feature data; acquiring ground elevation data of a preset actual measurement point in the forest region to be measured; and obtaining seasonal factors and vegetation type data of the forest region to be measured. Through multi-source remote sensing data collection, the technical bottleneck that traditional single optical remote sensing cannot penetrate through dense crowns is directly broken through, and areas with different canopy densities are covered.

The invention discloses a vegetation classification method based on space-time multi-modal deep learning, and relates to the technical field of image processing, and the method comprises the steps: obtaining a multi-temporal optical image, a radar image and digital elevation model data of a to-be-classified region, and forming multi-modal data; extracting spectral features, microwave features, topographic features and texture features to form fusion features; calculating graph node features, and updating the graph node features to form graph features; and fusing the fusion features and the graph features to obtain pixel-level coarse classification logs, carrying out feature extraction and fusion on the fusion features to form region-level coarse classification logs, carrying out fusion to obtain a coarse classification probability, and carrying out fine classification on coarse basic features to obtain a final classification result. The method provided by the invention can adapt to a mountainous area environment with multiple clouds, multiple shadows and large topographic relief, improves the stability and classification fineness of vegetation type identification, and is suitable for wide-range vegetation monitoring and ecological assessment scenes.

This invention discloses a coupled simulation and prediction system for regional water and carbon cycle processes, relating to the field of watershed carbon cycle simulation. The system includes a data acquisition module that determines the watershed extent and identifies riverbank areas within a unified grid system. Simultaneously, vegetation type and soil organic carbon are mapped to riverbank grid cells, consistently expressing the spatial location of carbon sources in relation to hydrological watershed migration. This facilitates subsequent clarification of carbon source distribution. Based on a DEM (Digital Elevation Model), flow direction and confluence paths are constructed. Runoff in each grid is calculated under meteorological conditions and gradually converges into the river channel, establishing hydrodynamic transport paths with clear upstream and downstream relationships. This provides a physical carrier for carbon transport. Based on soil organic carbon and vegetation type in the riverbank grids, the system quantifies the differences in carbon supply capacity of different riverbank areas during regional water migration. Riverbank carbon sources are traced along the confluence paths. Spatial weights are constructed using distance attenuation and connectivity, mapping carbon sources to each grid cell. Carbon flux is then coupled with runoff to calculate and accumulate through gradual transport.

This application discloses a vegetation classification method based on spatiotemporal multimodal deep learning, relating to the field of image processing technology. The method includes: acquiring multi-temporal optical images, radar images, and digital elevation model data of the region to be classified, forming multimodal data; extracting spectral features, microwave features, topographic features, and texture features to form fused features; calculating graph node features and updating the graph node features to form graph features; fusing the fused features and graph features to obtain pixel-level coarse classification logits; extracting and fusing the fused features to form region-level coarse classification logits; fusing to obtain coarse classification probabilities; and performing fine classification on the coarse class basic features to obtain the final classification result. The method of this application can adapt to mountainous environments with frequent cloud cover, shadows, and large topographic relief, improving the stability and classification precision of vegetation type identification, and is suitable for large-scale vegetation monitoring and ecological assessment scenarios.

The present application relates to the technical field of coal mining and discloses a method for magnetically controlling phreatic water level during coal mining. The method comprises: acquiring pre-mining information data; on the basis of a vegetation type and a soil sample from a phreatic layer, determining a vegetation growth range, the vegetation growth range comprising a water table depth range and a soil moisture content range; on the basis of the vegetation growth range and the pre-mining information data, determining the on or off state of a magnetizing device during mining; if the pre-mining water level of the phreatic aquifer is above the upper limit of a water table depth range, controlling the magnetizing device to be turned off and regulating corresponding information data; and if the post-mining water level of the phreatic aquifer is above the regulated water level of the phreatic aquifer, controlling the magnetizing device to be turned on, and regulating the magnetization intensity, so that post-mining information data is within the vegetation growth range. The present application effectively reduces the evaporation of groundwater and the occurrence of regional salinization, and can effectively protect the ecological environment and implement coal mining.

The application discloses a kind of highland surface vegetation coverage and carbon sink function remote sensing monitoring method, it is related to remote sensing ecological monitoring technical field, comprising the following steps: the highland remote sensing image is divided into multiple image blocks, and the multispectral feature vector of each image block is extracted;Image block is input into the spatial-semantic structure recognition model based on the structure constructed by Transform;Based on spatial dependence relationship and semantic association relationship, the structure recognition result and vegetation type classification result of image block are generated;According to structure recognition result and vegetation type classification result, the carbon absorption parameter corresponding to vegetation type and structure is combined, and the carbon sink function index of corresponding image area is calculated;The monitoring result including surface vegetation coverage distribution map, vegetation structure diagram and carbon sink function spatial distribution map is output.The application solves the problem of insufficient carbon sink function remote sensing evaluation precision under the complex structure environment of highland area.

The invention relates to the technical field of ecological remote sensing and intelligent restoration, in particular to an arbor, shrub and grass vegetation detection and analysis method and system based on an unmanned aerial vehicle, and the method comprises the steps: fusing the canopy height, texture and morphological features to achieve the fine classification of arbor, shrub and grass; combining the multi-temporal unmanned aerial vehicle image and the long-time sequence satellite data to construct a multi-dimensional feature vector, and generating a 1m-resolution vegetation degradation map through an optimized random forest model; and making a differential rejuvenation scheme according to the degradation grade, the vegetation type and the topographic condition. According to the method, the problems of vegetation misclassification, low degradation identification precision and lack of adaptability of rejuvenation measures can be effectively solved, and the precision and efficiency of ecological restoration are improved.

The application discloses a kind of grassland canopy height estimation methods based on multi-angle optical remote sensing data in Xilingol League, and relates to canopy height estimation technical field. Including: S1, data acquisition and pretreatment;S2, vegetation index calculation;S3, ground measured data matching;S4, model construction and inversion;S5, precision verification and optimization.The application is integrated by "multi-model+multi-index+multi-angle", effectively weaken the interference caused by the diversity of vegetation type, terrain undulation and soil background difference, improve the adaptability and robustness of model under complex spatial heterogeneity condition, at the same time, the application makes full use of multi-angle observation information, compensates the deficiency that single-angle method ignores multi-direction scattering and angle effect, can comprehensively reveal the three-dimensional structure characteristics of grassland canopy, further improve the precision and consistency of inversion result.

The invention discloses a forest and grass landscape pattern identification method based on unmanned aerial vehicle images, and the method comprises the steps: carrying out the definition and cloud shielding screening of images collected by an unmanned aerial vehicle through a random forest algorithm, and obtaining a high-quality first image data set; then, in combination with vegetation types and spatial distribution characteristics, landscape elements are classified by using a support vector machine, acquisition parameters are optimized, and a second image data set is generated. Further considering the influence of seasonal variation on spectral characteristics, selecting representative time phase images, and constructing a third image data set; and plaque information is calculated and fine plaques are fused to obtain fused plaque distribution data, and then multi-scale pattern indexes such as a shape index, an aggregation degree, a separation degree and a spreading degree are analyzed. And finally, indexes related to image definition are extracted, and a landscape pattern report is generated. According to the method, full-process automation from data screening, element classification, time phase selection to pattern analysis is realized, and the recognition precision and adaptability of the forest and grass landscape pattern are remarkably improved.

The invention relates to the technical field of image recognition, in particular to a vegetation coverage area remote sensing recognition method and system, and the method comprises the steps: obtaining a remote sensing image of a vegetation coverage area; obtaining the change complexity of each pixel in the remote sensing image under each wave band according to the change complexity of the reflectivity of each pixel in the remote sensing image under each wave band; the method comprises the following steps: acquiring an extreme value bit order dislocation degree and an extreme value dislocation degree between each pixel in a remote sensing image and each preset adjacent pixel in each wave band, further obtaining a similar extreme value dislocation degree of each pixel in the remote sensing image in each wave band, and performing region merging processing on the pixels in the remote sensing image in each wave band to obtain a merged image in each wave band; acquiring the change recognition degree of each pixel in the combined image; and classifying the pixels in the remote sensing image, and identifying the types of vegetation in each type. The invention aims to improve the accuracy of identifying the vegetation types in the vegetation coverage area.

The invention relates to a sand arbor, shrub, grass and sand identification and segmentation method, system and device and a medium, and the method comprises the steps: constructing a classification and identification model based on a U-Net model, and constructing a training data set for model training based on an unmanned plane remote sensing image of a sand sparse vegetation sample land; performing classification identification detection on the to-be-identified image data based on the classification identification model to obtain a vegetation segmentation result; and based on the vegetation segmentation result, determining the coverage and density of different vegetation types on the sand sparse vegetation sample land through spatial intersection analysis. The method can be widely applied to the technical field of arbor-shrub-grass sand recognition.

The application provides a carbon sink potential prediction method and device fusing remote sensing and model technology. The carbon sink potential prediction method fusing remote sensing and model technology comprises the following steps: inputting multi-temporal remote sensing data into a vegetation type identification model based on historical remote sensing data and historical vegetation type labels to obtain vegetation structure dynamic distribution data; inputting the vegetation structure dynamic distribution data into a regional carbon-nitrogen cycle model based on historical crop distribution data and vegetation and soil dynamic data to obtain a regional carbon-nitrogen cycle dynamic simulation result; and determining carbon sink data according to the regional carbon-nitrogen cycle dynamic simulation result. The application can break through the bottleneck of insufficient regional scale carbon accounting data and high uncertainty, and realize carbon sink maximization.

This invention provides a method, system, device, and medium for identifying vegetation lodging risk around power facilities, relating to the field of power facility risk identification. The method includes: extracting vegetation canopy data of the transmission corridor of the target power facility from satellite multispectral images; calculating the initial tilt angle and lodging direction; obtaining a list of potential lodging areas; acquiring visible light images of the transmission corridor using ground image acquisition equipment; extracting vegetation pixels and obtaining satellite normalized vegetation index values ​​for corresponding locations at multiple time phases to construct a seasonal fluctuation feature sequence; extracting texture features from corresponding locations in the ground visible light image; constructing a fused feature vector and inputting it into a convolutional neural network for vegetation type classification; and calculating the risk identification result for each vegetation pixel by combining the initial tilt angle and lodging direction. This invention improves the accuracy of vegetation lodging risk detection around power facilities.

This invention relates to a method for estimating surface dead combustible load in subtropical forests using multi-source remote sensing and machine learning. The method includes: acquiring and preprocessing multi-source remote sensing data and auxiliary data within a target area; acquiring preprocessed multi-source remote sensing data and auxiliary data; extracting features from the preprocessed multi-source remote sensing data and auxiliary data to obtain feature variables; selecting variables based on the feature variables and measured surface dead combustible load components; constructing a combustible load prediction model using multiple linear regression and machine learning models; evaluating the combustible load prediction model to obtain the optimal model; and using the optimal model to perform regional inversion of surface dead combustible load for all vegetation types within the target area to obtain spatial distribution data of surface dead combustible load. This invention combines multi-source remote sensing with machine learning to provide methodological support for fire risk assessment and precise combustible management in subtropical forests.