Sky-ground integrated phenological phenomenon monitoring system and method

Abstract

The invention relates to a sky-ground integrated phenological phenomenon monitoring system and method. The system comprises a data acquisition module, a data processing module, a phenological change analysis module and an early warning module. The data acquisition module is used for acquiring multi-source data of vegetation phenology; the data processing module is used for fusing the multi-source data; the phenological change analysis module is used for predicting the phenological change in the future time by adopting a machine learning method based on the fused data; and the early warning module is used for sending early warning information to a user when an abnormal phenological event is identified according to the prediction result. The vegetation phenological change can be effectively monitored through various data acquisition means, accurate phenological prediction and early warning are provided through data fusion and intelligent analysis, and then support is provided for ecological monitoring, agricultural management and climate change research.

Description

Technical Field

[0001] The present invention relates to the fields of meteorological monitoring and environmental science and technology, and in particular to a sky-ground integrated phenological phenomenon monitoring system and method. Background Art

[0002] Phenology studies the growth, development, and key life cycle events of plants and animals, such as flowering, fruiting, and leaf color changes. The impact of climate change on phenology is becoming increasingly significant, so timely monitoring of phenological changes is crucial for understanding ecosystems, agricultural production, and the impacts of climate change.

[0003] Traditional phenological monitoring relies primarily on ground-based observation sites. This approach is limited by spatial distribution, monitoring accuracy, and data timeliness, and cannot meet the needs of large-scale, multi-dimensional, and highly timely monitoring. Satellite remote sensing technology provides extensive spatial coverage, but its data timeliness is poor and it struggles to capture detailed information about local areas. Unmanned aerial vehicle (UAV) technology can provide high-resolution local monitoring data, but still faces challenges with limited flight time and coverage area. Ground-based sensors (such as infrared cameras) can provide detailed real-time data but are limited by location.

[0004] Therefore, how to integrate different monitoring methods, make up for their respective shortcomings, and provide a high-precision, high-timeliness, large-scale phenological monitoring system is an important challenge in the current field of phenological monitoring. Summary of the Invention

[0005] The present invention provides a sky-ground integrated vegetation phenology monitoring system and method. The system can effectively monitor vegetation phenology changes through multiple data collection methods (including satellite remote sensing, drone photography, ground phenology cameras, and manual recording), and provide accurate phenology predictions and early warnings through data fusion and intelligent analysis, thereby providing support for ecological monitoring, agricultural management, and climate change research.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] An integrated sky-ground phenological phenomenon monitoring system includes: a data acquisition module, a data processing module, a phenological change analysis module, and an early warning module;

[0008] The data acquisition module is used to collect multi-source data on vegetation phenology;

[0009] The data processing module is used to fuse the multi-source data;

[0010] The phenological change analysis module is used to predict phenological changes in the future using a machine learning method based on the fused data;

[0011] The early warning module is used to send early warning information to the user when an abnormal phenological event is identified in the prediction results.

[0012] Optionally, the data acquisition module includes: a first acquisition unit, a second acquisition unit, a third acquisition unit, and a fourth acquisition unit;

[0013] The first acquisition unit is used to collect remote sensing data of vegetation; wherein the remote sensing data includes: vegetation index, chlorophyll content, surface temperature change, and vegetation coverage;

[0014] The second collection unit is used to collect plant growth status data and environmental data from the air;

[0015] The third collection unit is used to collect plant physiological change data on the ground and identify plant phenological events;

[0016] The fourth collection unit is used to manually record the phenological events of plants.

[0017] Optionally, the data processing module includes: a pre-processing unit and a fusion unit;

[0018] The pre-processing unit is used to remove noise, fill missing data, and correct outliers on the multi-source data;

[0019] The fusion unit is used to fuse the preprocessed data through spatial, temporal and semantic fusion methods to generate a vegetation phenology dataset.

[0020] Optionally, the phenological change analysis module includes: a feature extraction unit, a model training unit, and a prediction unit;

[0021] The feature extraction unit is used to extract phenological features of vegetation phenology from the fused data; wherein the phenological features are used to reflect the phenological phenomena of plants at different stages;

[0022] The model training unit is used to train the machine learning model based on the phenological characteristics to obtain a prediction model;

[0023] The prediction unit is used to predict the phenological changes of vegetation in the future using the prediction model.

[0024] The present invention also proposes a method for monitoring phenological phenomena in an integrated sky-ground manner, the method comprising:

[0025] Collect multi-source data on vegetation phenology;

[0026] fusing the multi-source data;

[0027] Based on the fused data, machine learning methods are used to predict future phenological changes;

[0028] When abnormal phenological events are identified in the prediction results, early warning information is sent to users.

[0029] Optionally, multi-source data on vegetation phenology may be collected, including:

[0030] Collecting remote sensing data of vegetation; wherein the remote sensing data includes: vegetation index, chlorophyll content, surface temperature change, and vegetation coverage;

[0031] Aerial collection of plant growth status data and environmental data;

[0032] Collect data on plant physiological changes on the ground and identify plant phenological events;

[0033] Manually record plant phenological events.

[0034] Optionally, fusing the multi-source data includes:

[0035] Preprocessing the multi-source data; wherein the preprocessing includes: noise removal, missing data filling, and outlier correction;

[0036] The preprocessed data are fused through spatial, temporal and semantic fusion methods to generate a vegetation phenology dataset.

[0037] Optionally, using machine learning methods to predict future phenological changes includes:

[0038] Extracting phenological characteristics of vegetation phenology from the fused data; wherein the phenological characteristics are used to reflect the phenological phenomena of plants at different stages;

[0039] Based on the phenological characteristics, a machine learning model is trained to obtain a prediction model;

[0040] The prediction model is used to predict the phenological changes of vegetation in the future.

[0041] The beneficial effects of the present invention are:

[0042] The present invention collects multi-source data on vegetation phenology through a data acquisition module to effectively monitor vegetation phenological changes; fuses the multi-source data through a data processing module; uses a phenological change analysis module to predict future phenological changes based on the fused data using machine learning methods; and sends warning information to users when an abnormal phenological event is identified in the prediction results through an early warning module. The present invention provides accurate phenological predictions and early warnings through data fusion and intelligent analysis, thereby providing support for ecological monitoring, agricultural management, and climate change research. The present invention can achieve large-scale, real-time, and high-precision monitoring of phenological phenomena and can be widely used in multiple fields such as climate change, ecological restoration, and agricultural management. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 The present invention provides a flow chart of a system for monitoring phenological phenomena in an integrated sky-ground manner. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] like Figure 1 As shown, this embodiment proposes a sky-ground integrated phenological phenomenon monitoring system, including: a data acquisition module, a data processing module, a phenological change analysis module, and an early warning module;

[0048] The data acquisition module is used to collect multi-source data on vegetation phenology; the data processing module is used to fuse multi-source data; the phenological change analysis module is used to predict future phenological changes based on the fused data using machine learning methods; and the early warning module is used to send early warning information to users when abnormal phenological events are identified in the prediction results.

[0049] Furthermore, the data acquisition module includes: a first acquisition unit, a second acquisition unit, a third acquisition unit, and a fourth acquisition unit; the first acquisition unit is used to collect remote sensing data of vegetation; wherein the remote sensing data includes: vegetation index, chlorophyll content, surface temperature change, and vegetation coverage; the second acquisition unit is used to collect plant growth status data and environmental data from the air; the third acquisition unit is used to collect plant physiological change data on the ground and identify plant phenological events; the fourth acquisition unit is used to manually record plant phenological events.

[0050] In this embodiment, the first acquisition unit uses satellite remote sensing technology to conduct remote sensing observations of vegetation in a large area through satellites equipped with multispectral sensors, providing support for the macro trend analysis of phenological phenomena. The data collected by the satellite mainly include: Vegetation index (such as NDVI): reflecting the growth status of vegetation. Chlorophyll content: reflecting the photosynthetic capacity of plants. Surface temperature changes: monitoring the thermal response of vegetation through data in the thermal infrared band. Vegetation coverage: reflecting the vegetation coverage of the area.

[0051] Specifically, the first acquisition unit uses existing high-resolution satellite remote sensing data (such as Sentinel-2 and Landsat) as the primary data source, acquiring data on vegetation cover, vegetation indices (such as NDVI), and chlorophyll content over a large area through satellite imagery. Conventional remote sensing processing methods, including radiation correction, atmospheric correction, and geometric correction, are used to improve the accuracy and consistency of remote sensing data. To increase the frequency of monitoring, long-term remote sensing data is combined with image time series analysis to detect phenological changes in vegetation.

[0052] In this embodiment, the second acquisition unit adopts drone monitoring. The drone is equipped with a high-definition RGB camera and environmental sensors (temperature, humidity, soil moisture sensors, etc.), which can perform high-resolution and fine monitoring of specific areas, capture data such as vegetation growth conditions and phenological changes, and is especially suitable for fine monitoring in a small range. Mainly including: Plant growth status images: High-resolution images help monitor vegetation leaf changes, health status, etc. Environmental data: Environmental information such as temperature, humidity, and soil moisture is collected through the onboard sensors. The drone flight adopts automated route planning and real-time monitoring system to ensure coverage of the target area while reducing errors caused by human operation.

[0053] In this embodiment, the third acquisition unit uses a ground-based phenological camera. The ground-based phenological camera uses high-precision sensors to capture physiological changes in plants, such as leaf temperature and humidity, in real time, and provides accurate records of plant phenological phenomena through long-term continuous monitoring. This module can automatically identify important phenological events such as flowering, fruiting, and leaf changes in plants. It mainly includes: fixed-point plant photography: able to identify important phenological events such as flowering, fruiting, and leaf changes. Leaf temperature: reflects the water status and physiological activities of the plant. Ambient temperature and humidity: monitor the impact of environmental changes on plants.

[0054] In this embodiment, the fourth data collection unit uses manual recording, with field staff manually recording plant phenological events based on intuitive observation, including flowering, fruiting, leaf color change, etc. These records supplement and verify the automated monitoring data, ensuring data accuracy and comprehensiveness.

[0055] Furthermore, the data processing module includes: a preprocessing unit and a fusion unit; the preprocessing unit is used to denoise, fill missing data, and correct outliers for multi-source data; the fusion unit is used to fuse the preprocessed data through spatial, temporal and semantic fusion methods to generate a vegetation phenology dataset.

[0056] In this embodiment, to accurately monitor vegetation phenology, the data processing module uses multi-source data fusion technology. By integrating and analyzing satellite remote sensing data, drone data, ground-based phenology camera data, and manual recording data, the system can generate a comprehensive vegetation phenology monitoring dataset.

[0057] Preprocessing unit: Preprocesses satellite remote sensing, drone imagery, infrared camera and manually recorded data, including noise removal, missing data filling, outlier correction, etc., to ensure data accuracy and consistency.

[0058] Fusion Unit: This unit integrates data from different monitoring sources through spatial, temporal, and semantic fusion techniques. This fusion generates a comprehensive vegetation phenology dataset that users can view and analyze through a web platform.

[0059] Furthermore, the phenological change analysis module includes: a feature extraction unit, a model training unit, and a prediction unit;

[0060] The feature extraction unit is used to extract the phenological characteristics of vegetation phenology from the fused data; wherein the phenological characteristics are used to reflect the phenological phenomena of plants at different stages;

[0061] Model training unit, used to train the machine learning model based on phenological characteristics to obtain a prediction model;

[0062] The prediction unit is used to predict the phenological changes of vegetation in the future using a prediction model.

[0063] Specifically, in this embodiment, the phenological change analysis module uses machine learning and deep learning algorithms (such as random forests, support vector machines, and neural networks) to intelligently analyze phenological phenomena and extract patterns in vegetation phenological changes. By establishing a predictive model, the system can predict phenological changes over a period of time, such as flowering period and fruit ripening period.

[0064] The system not only monitors vegetation phenological changes in real time but also provides forecasts and early warnings based on historical data and meteorological factors. By learning from historical data and analyzing trends, the system can predict future phenological changes and issue early warning notifications before unusual phenological events (such as frost and drought) occur.

[0065] (1) Phenological event prediction

[0066] The system uses regression analysis and classification models, combined with meteorological data, to predict the key growth stages of vegetation, helping users to grasp the changing trends of vegetation in advance.

[0067] (2) Warning Notice

[0068] When the system identifies potential abnormal phenological events, the early warning module will automatically send warning information to the user, along with a response plan, to ensure that the user can take measures in advance to deal with possible ecological disasters.

[0069] This embodiment presents monitoring data and analysis results to users through a web platform. The platform adopts a responsive design and supports access from multiple terminals such as PC and mobile terminals. Users can view monitoring data in real time, generate reports, obtain forecast results, and set warning parameters.

[0070] (1) Data display:

[0071] The system displays remote sensing data, drone data and ground phenological camera data in the form of charts, maps, etc., allowing users to intuitively view changes in vegetation phenology.

[0072] (2) Intelligent Analysis Report:

[0073] Users can generate analysis reports through the web platform to understand the trends and potential impacts of phenological changes. The system also provides phenological forecasts and early warning reports to help users prepare for response.

[0074] (3) User management and authority control:

[0075] The system provides different access rights for users with different roles (such as administrators, researchers, managers, and data collectors) to ensure data security and efficient operation of the platform.

[0076] The system implementation process of this embodiment is as follows:

[0077] (1) Data collection and upload

[0078] Each monitoring device (satellite, drone, ground infrared camera, manual recorder) collects data according to the predetermined plan and uploads the collected data to the web platform in real time.

[0079] (2) Data preprocessing and cleaning

[0080] The platform first cleans and preprocesses the uploaded data to handle missing data, remove noise, and correct abnormal data.

[0081] (3) Data fusion and analysis

[0082] After processing, the data is integrated to generate a comprehensive vegetation phenology monitoring dataset. The platform then uses intelligent analysis algorithms to analyze vegetation phenology changes and generate forecast and early warning reports.

[0083] (4) Results display and sharing

[0084] Through the web platform, users can view and analyze monitoring data, generate reports, and obtain early warning information. All data and analysis results can be shared with team members or external researchers.

[0085] (5) Warning notification and response

[0086] If the system identifies potential abnormal phenological events (such as frost, drought, etc.), it will automatically send an early warning notification to the user, along with recommended response measures.

[0087] This embodiment has the following technical advantages:

[0088] (1) Multi-source data fusion and high-precision monitoring

[0089] Through the fusion of multi-source data from satellite remote sensing, UAVs, ground phenological cameras and manual records, the present invention can achieve high-precision vegetation phenology monitoring in large areas and small local areas.

[0090] (2) Intelligent prediction and early warning

[0091] Based on big data analysis and machine learning technology, the system can accurately predict changes in vegetation phenology and issue early warnings to help users prepare in advance.

[0092] (3) Convenience and practicality of the Web platform

[0093] Users can view vegetation phenology change data, generate reports, and receive early warning notifications anytime and anywhere through the web platform. The operation is simple and easy to share.

[0094] (4) Data security and collaboration functions

[0095] The system has a complete data security mechanism and user authority management, which can effectively ensure data security and support team collaboration.

[0096] Example 2

[0097] This embodiment proposes a method for monitoring phenological phenomena using integrated sky and ground systems, including:

[0098] Collect multi-source data on vegetation phenology;

[0099] Fusion of multi-source data;

[0100] Based on the fused data, machine learning methods are used to predict future phenological changes;

[0101] When abnormal phenological events are identified in the prediction results, early warning information is sent to users.

[0102] Furthermore, the multi-source data of vegetation phenology are collected including:

[0103] Collect remote sensing data of vegetation; remote sensing data includes vegetation index, chlorophyll content, surface temperature change, and vegetation coverage;

[0104] Aerial collection of plant growth status data and environmental data;

[0105] Collect data on plant physiological changes on the ground and identify plant phenological events;

[0106] Manually record plant phenological events.

[0107] Specifically, in this embodiment, the data collection technology adopts:

[0108] (1) Satellite remote sensing technology

[0109] The system uses existing high-resolution satellite remote sensing data (such as Sentinel-2 and Landsat) as its primary data source, acquiring data on vegetation cover, vegetation indices (such as NDVI), and chlorophyll content over large areas through satellite imagery. Conventional remote sensing processing methods, including radiometric correction, atmospheric correction, and geometric correction, are employed to improve the accuracy and consistency of remote sensing data. To increase monitoring frequency, long-term remote sensing data is combined with image time series analysis to detect phenological changes in vegetation.

[0110] (2) UAV filming technology:

[0111] The drone is equipped with high-resolution cameras (RGB and multispectral) and environmental sensors (temperature, humidity, soil moisture, etc.), enabling precise local vegetation measurements. The drone's flight utilizes automated route planning and real-time monitoring systems to ensure coverage of the target area while minimizing human error.

[0112] (3) Ground-based infrared camera technology

[0113] High-precision infrared cameras are used to capture real-time data such as leaf temperature and humidity. Infrared imaging technology efficiently and contactlessly captures plant surface temperature and heat transfer characteristics, helping to monitor plant physiological status and moisture levels. This technology is invaluable for monitoring plant responses throughout the growing season.

[0114] (4) Manual recording technology

[0115] Manual recording primarily involves field staff using visual observation and traditional recording methods to annotate phenological events. Based on the actual growth of vegetation on site (such as flowering and maturity, leaf color changes, etc.), staff enter standardized data and upload it to the system for archiving. This process complements automated monitoring and ensures the comprehensiveness and authenticity of the data.

[0116] Furthermore, the fusion of multi-source data includes:

[0117] Preprocess multi-source data; preprocessing includes: noise removal, missing data filling, and outlier correction;

[0118] The preprocessed data are fused through spatial, temporal and semantic fusion methods to generate a vegetation phenology dataset.

[0119] Specifically, in this embodiment, the data fusion and analysis technology:

[0120] Multi-source data fusion technology:

[0121] Data fusion is one of the key technologies of this invention, primarily involving the integration of satellite remote sensing data, drone data, ground-based infrared camera data, and manually recorded data. During the fusion process, each data source is first aligned in time and space to ensure data consistency and comparability. Then, through methods such as spatial interpolation, time series analysis, and pattern recognition, data from these different sources are combined to generate a comprehensive dataset on vegetation phenology. The innovation of this fusion technology lies in its ability to compensate for the limited accuracy of remote sensing data by combining it with high-resolution drone data, thereby improving the precision of phenological monitoring.

[0122] Data cleaning and preprocessing technology:

[0123] Since the data quality from different monitoring sources may vary, the system will first clean and pre-process the collected data. Specifically, it includes:

[0124] Denoising: Use statistical methods or filtering algorithms to remove noise from data and improve data quality.

[0125] Missing data filling: Use interpolation or machine learning models to fill in missing data to ensure data integrity.

[0126] Outlier detection and correction: By analyzing data distribution patterns, abnormal data can be identified and corrected to ensure data accuracy.

[0127] Furthermore, machine learning methods are used to predict future phenological changes, including:

[0128] Extract the phenological characteristics of vegetation phenology from the fused data; wherein, the phenological characteristics are used to reflect the phenological phenomena of plants at different stages;

[0129] Based on phenological characteristics, the machine learning model is trained to obtain a prediction model;

[0130] Use prediction models to predict future phenological changes in vegetation.

[0131] Specifically, in this embodiment, machine learning and deep learning algorithms (such as random forests, support vector machines, neural networks, etc.) are used to perform intelligent analysis on the fused data. The algorithm can identify the laws of vegetation phenology, infer the growth stage of vegetation, the impact of climate change on plants, future phenological trends, etc. For example, by analyzing the time series data of vegetation index (NDVI) and leaf temperature, key phenological information such as flowering period and fruit ripening period can be extracted. In addition, this embodiment predicts future vegetation phenology through regression analysis or classification models to provide users with early warning information.

[0132] This embodiment combines machine learning models with data time series analysis technology to predict vegetation phenological changes. Based on historical data and meteorological conditions, the model can predict key vegetation growth stages and issue early warning notifications before unusual phenological events (such as frost and drought) occur. This automated and intelligent warning system provides users with timely response plans.

[0133] Web platform design and implementation technology of this embodiment

[0134] (1) Front-end technology

[0135] The web platform's front-end interface is implemented using technologies such as HTML5, CSS3, and JavaScript. Modern front-end frameworks (such as React and Vue) enable a responsive system design, allowing the platform to run smoothly on various devices (PCs, mobile phones, and tablets). Users can intuitively view monitoring data, generate reports, and set alerts through a graphical interface. Charts and maps utilize visualization frameworks such as ECharts and Leaflet, enabling a multi-dimensional display of monitoring data.

[0136] (2) Backend technology

[0137] The web platform's backend, developed in programming languages ​​such as Python, Node.js, and Java, is responsible for processing user requests, invoking algorithm modules, and accessing database data. The backend interacts with the frontend via a RESTful API, ensuring real-time data updates and display. The backend uses machine learning libraries for data analysis and modeling, and also generates future phenological trends through predictive models.

[0138] (3) Database technology

[0139] The system uses a combination of relational databases (such as MySQL and PostgreSQL) and non-relational databases (such as MongoDB) for data storage. Remote sensing imagery, drone imagery, and ground sensor data are stored in a distributed database, supporting large-scale data storage and query. All monitoring data is automatically backed up to ensure data security and reliability.

[0140] (4) Data security and user rights management:

[0141] The system has a strict user rights management mechanism, providing different data access permissions to different users based on their roles (e.g., administrator, data analyst, ecologist, etc.). All user actions are logged to ensure data security. HTTPS encryption is used for data transmission to ensure data security during transmission.

[0142] To validate the effectiveness of the proposed system approach, a ground-based, integrated sky-ground vegetation phenology monitoring system was deployed and applied in a national nature reserve in southern China. The area boasts typical subtropical evergreen broad-leaved forests and wetland vegetation, with distinct phenological seasonality, making it a suitable testing area for the system.

[0143] (1) Monitoring equipment deployment

[0144] Satellite remote sensing data acquisition

[0145] Sentinel-2 images (10-meter resolution, 5-day revisit period) are used as the main remote sensing data source, and vegetation indices such as NDVI, EVI, and NDWI are regularly captured through the Google Earth Engine platform. The time span is from January 2022 to December 2023, with a frequency of once every 10 days.

[0146] UAV aerial photography system configuration

[0147] Three representative plots (A, B, and C) were set up within the reserve and monthly flights were conducted using drones equipped with RGB and multispectral cameras (such as the Micasense RedEdge). The flights were set at an altitude of 120 meters, covering an area of ​​approximately 50 hectares per flight, and obtaining images with a resolution better than 10 cm.

[0148] Ground infrared / visible light camera installation

[0149] Each plot is equipped with a PhenoCam infrared imaging camera, mounted atop an observation tower and oriented toward a typical forest canopy. Images are captured at a rate of one per hour, automatically uploaded to an edge server, and synchronized to a web platform via a 4G network.

[0150] Manual observation and recording mechanism

[0151] In the middle of each month, staff at the ecological station manually record the phenological stages (germination, leaf expansion, flowering, leaf fall, etc.) of representative plant species in the sample area in accordance with the "China Vegetation Phenology Observation Manual" and upload them to the web system via mobile phones.

[0152] (2) Data collection and fusion processing

[0153] Remote sensing data preprocessing

[0154] Sen2Cor was used for radiometric and atmospheric correction, and NDVI calculations were performed using GEE. The images were uniformly resampled to 10 meters and cloud masked.

[0155] Fusion of drone and ground data

[0156] The high-resolution images acquired by the drone are registered to the remote sensing image grid, and the same indicators as remote sensing (NDVI, GNDVI, and red edge vegetation index) are extracted to improve accuracy. The infrared image and drone data are spatially matched and superimposed to analyze leaf temperature changes.

[0157] Time Series Analysis

[0158] Time series curves were constructed for all monitored indicators (NDVI, leaf temperature, humidity, etc.) to identify key phenological turning points. An LSTM neural network was trained on data from different plots to extract the timing of phenological events (such as the start of leaf expansion and the duration of flowering).

[0159] (3) Implementation of prediction and early warning systems

[0160] In February 2023, by analyzing vegetation indices and temperature trends, the system successfully predicted that the red nanmu trees in plot A would enter their flowering phase in early March and sent an early warning message to researchers via the platform. Actual manual verification showed a prediction error of less than two days, confirming the model's effectiveness.

[0161] In May 2023, the system detected continued high temperatures and a downward trend in NDVI, and predicted that premature leaf fall would occur in sample area B, indicating possible drought stress. This was confirmed by manual verification and an early warning was issued 10 days in advance, which has significant ecological management value.

[0162] (4) Web platform operation and user feedback

[0163] The platform displays to users through the browser interface: multi-temporal and spatial NDVI curves; phenological event timeline.

[0164] This embodiment verifies the full-process capabilities of the system of the present invention in "sky-ground" data fusion, phenological analysis, early warning prediction, and data visualization in real scenarios, and has good practical application prospects.

[0165] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A sky-ground integrated phenological phenomenon monitoring system, characterized in that: include: Data acquisition module, data processing module, phenological change analysis module, and early warning module; The data acquisition module is used to collect multi-source data on vegetation phenology; The data processing module is used to fuse the multi-source data; The phenological change analysis module is used to predict phenological changes in the future using a machine learning method based on the fused data; The early warning module is used to send early warning information to the user when an abnormal phenological event is identified in the prediction results.

2. The sky-ground integrated phenological phenomenon monitoring system according to claim 1, characterized in that: The data acquisition module includes: a first acquisition unit, a second acquisition unit, a third acquisition unit, and a fourth acquisition unit; The first acquisition unit is used to collect remote sensing data of vegetation; wherein the remote sensing data includes: vegetation index, chlorophyll content, surface temperature change, and vegetation coverage; The second collection unit is used to collect plant growth status data and environmental data from the air; The third collection unit is used to collect plant physiological change data on the ground and identify plant phenological events; The fourth collection unit is used to manually record the phenological events of plants.

3. The sky-ground integrated phenological phenomenon monitoring system according to claim 1, characterized in that: The data processing module includes: a pre-processing unit and a fusion unit; The pre-processing unit is used to remove noise, fill missing data, and correct outliers on the multi-source data; The fusion unit is used to fuse the preprocessed data through spatial, temporal and semantic fusion methods to generate a vegetation phenology dataset.

4. The sky-ground integrated phenological phenomenon monitoring system according to claim 1, characterized in that: The phenological change analysis module includes: a feature extraction unit, a model training unit, and a prediction unit; The feature extraction unit is used to extract phenological features of vegetation phenology from the fused data; wherein the phenological features are used to reflect the phenological phenomena of plants at different stages; The model training unit is used to train the machine learning model based on the phenological characteristics to obtain a prediction model; The prediction unit is used to predict the phenological changes of vegetation in the future using the prediction model.

5. A method for monitoring phenological phenomena by integrating air and ground, characterized in that: The method applied to the sky-ground integrated phenological phenomenon monitoring system according to any one of claims 1 to 4 comprises: Collect multi-source data on vegetation phenology; fusing the multi-source data; Based on the fused data, machine learning methods are used to predict future phenological changes; When abnormal phenological events are identified in the prediction results, early warning information is sent to users.

6. The sky-ground integrated phenological phenomenon monitoring method according to claim 5, characterized in that: The multi-source data for collecting vegetation phenology include: Collecting remote sensing data of vegetation; wherein the remote sensing data includes: vegetation index, chlorophyll content, surface temperature change, and vegetation coverage; Aerial collection of plant growth status data and environmental data; Collect data on plant physiological changes on the ground and identify plant phenological events; Manually record plant phenological events.

7. The sky-ground integrated phenological phenomenon monitoring method according to claim 5, characterized in that: Fusing the multi-source data includes: Preprocessing the multi-source data; wherein the preprocessing includes: noise removal, missing data filling, and outlier correction; The preprocessed data are fused through spatial, temporal and semantic fusion methods to generate a vegetation phenology dataset.

8. The sky-ground integrated phenological phenomenon monitoring method according to claim 5, characterized in that: The use of machine learning methods to predict future phenological changes includes: Extracting phenological characteristics of vegetation phenology from the fused data; wherein the phenological characteristics are used to reflect the phenological phenomena of plants at different stages; Based on the phenological characteristics, a machine learning model is trained to obtain a prediction model; The prediction model is used to predict the phenological changes of vegetation in the future.

Images