Ecological meteorology and satellite remote sensing combined dynamic environment monitoring system

Abstract

The invention provides an ecological meteorological and satellite remote sensing combined environment dynamic monitoring system, and relates to the technical field of environment monitoring. An ecological meteorological monitoring data set is obtained through an ecological meteorological monitoring module; the satellite remote sensing monitoring module obtains a satellite remote sensing monitoring data set; the data fusion module fuses the ecological meteorological monitoring data set and the satellite remote sensing monitoring data set to generate an environment dynamic monitoring result; a risk prediction computing power analysis module performs risk prediction computing power analysis according to the environment dynamic monitoring result to determine environment risk prediction computing power; the computing power-driven risk prediction module drives the environment risk prediction channel to perform environment risk prediction to obtain an environment risk prediction coefficient; and the monitoring configuration adjustment module optimizes the configuration of the ecological meteorological monitoring module and the satellite remote sensing monitoring module according to the environmental risk prediction coefficient. The technical problems that an existing method is single in data source and insufficient in environment monitoring precision are solved, and the accuracy and reliability of dynamic environment monitoring and risk prediction are improved.

Description

Technical Field

[0001] The present application relates to the field of environmental monitoring technology, and in particular to an environmental dynamic monitoring system combining ecological meteorology and satellite remote sensing. Background Art

[0002] Environmental monitoring can provide timely data on ecological meteorology, pollution conditions, and ecological health status, and provide scientific research and decision-making support for ecological protection, climate change response, disaster warning, etc.

[0003] At present, the main environmental monitoring methods include ground-based ecological meteorological monitoring and remote sensing monitoring technology. The ground-based meteorological monitoring method obtains regional meteorological information through meteorological stations and sensors, which has the advantage of strong real-time performance, but its coverage is limited and it is difficult to comprehensively monitor ecological changes in a large area; satellite remote sensing technology obtains a wide range of environmental data through satellites or high-altitude platforms, such as surface cover, vegetation index, atmospheric composition, etc., which can achieve periodic monitoring of a wide area and provide large-scale environmental information, but the satellite imaging frequency is low and the accuracy is limited, which cannot meet the needs of high-timeliness and refined dynamic environmental monitoring. Most of these monitoring methods operate independently, lack an effective integration mechanism, and the data is isolated, making it difficult to accurately assess environmental risks. In addition, in terms of risk prediction, the existing monitoring technologies are mostly based on simple statistics or single indicator threshold judgments, lacking the ability to accurately predict by comprehensively considering multiple factors and their interrelationships, and at the same time, the monitoring configuration is fixed and difficult to flexibly adjust according to actual environmental risks. Summary of the invention

[0004] This application provides an environmental dynamic monitoring system that combines ecological meteorology and satellite remote sensing, which solves the technical problem that the existing environmental monitoring methods have a single data source, resulting in insufficient environmental monitoring accuracy, and achieves the technical effect of improving the accuracy and reliability of environmental dynamic monitoring and risk prediction, as well as the rationality of monitoring resource allocation.

[0005] In view of the above problems, the present application provides an environmental dynamic monitoring system combining ecological meteorology and satellite remote sensing, the system comprising: an ecological meteorology monitoring module, the ecological meteorology monitoring module is used to perform real-time monitoring of the target area and obtain an ecological meteorology monitoring data set; a satellite remote sensing monitoring module, the satellite remote sensing monitoring module is used to obtain a satellite remote sensing monitoring data set of the target area; a data fusion module, the data fusion module is used to perform data fusion on the ecological meteorology monitoring data set and the satellite remote sensing monitoring data set to generate an environmental dynamic monitoring result; a risk prediction computing power analysis module, the risk prediction computing power analysis module is used to analyze the environmental abnormality depth threshold according to the environmental abnormality depth threshold. The dynamic monitoring results are used to perform risk prediction computing power analysis on the target area to determine the environmental risk prediction computing power; a computing power-driven risk prediction module is used to drive the environmental risk prediction channel to perform environmental risk prediction on the target area based on the environmental risk prediction computing power to obtain an environmental risk prediction coefficient; a monitoring configuration adjustment module is used to optimize the monitoring configuration of the ecological meteorological monitoring module and the satellite remote sensing monitoring module based on the first environmental risk threshold and the second environmental risk threshold of the target area according to the environmental risk prediction coefficient, wherein the first environmental risk threshold is greater than the second environmental risk threshold.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The ecological meteorological monitoring module monitors the ecological meteorological elements of the target area in real time, obtains the ecological meteorological monitoring data set, and provides real-time ground ecological environment information. The satellite remote sensing monitoring module obtains the macro-environmental information of the target area, generates a satellite remote sensing monitoring data set, and provides macro observation information of the target area from a spatial perspective. The data fusion module fuses the ecological meteorological monitoring data set and the satellite remote sensing monitoring data set to obtain environmental dynamic monitoring results covering timeliness and spatial accuracy, providing more comprehensive and accurate data support for subsequent risk prediction. The risk prediction computing power analysis module analyzes the risk prediction computing power of the target area based on the environmental anomaly depth threshold and the environmental dynamic monitoring results, matches appropriate computing power resources, and determines the environmental risk prediction computing power. The computing power-driven risk prediction module drives the environmental risk prediction channel to predict the environmental risk of the target area based on the environmental risk prediction computing power, and converts complex environmental data into quantifiable environmental risk prediction coefficients, thereby more accurately reflecting the environmental risk status and providing an important basis for environmental management decision-making. The monitoring configuration adjustment module optimizes the monitoring configuration of the ecological meteorological monitoring module and the satellite remote sensing monitoring module based on the first environmental risk threshold and the second environmental risk threshold of the target area according to the environmental risk prediction coefficient, thereby ensuring that the monitoring system can adjust its monitoring strategy according to the current environmental risk level and achieve reasonable allocation of monitoring resources.

[0008] To summarize, this application achieves high-precision dynamic environmental monitoring and accurate risk assessment of the target area through the integration of ecological meteorological data and satellite remote sensing data, combined with computing power-driven risk prediction and dynamic configuration adjustment, overcomes the temporal and spatial limitations of a single monitoring method, significantly improves the comprehensiveness and accuracy of environmental monitoring, and at the same time improves the efficiency of monitoring resource utilization.

[0009] 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

[0010] Figure 1 A schematic diagram of the structure of the environmental dynamic monitoring system combining ecological meteorology and satellite remote sensing provided in an embodiment of the present application.

[0011] Figure 2 A schematic diagram of the process of obtaining environmental risk prediction computing power in the environmental dynamic monitoring system that combines ecological meteorology and satellite remote sensing provided in an embodiment of the present application.

[0012] Figure 3A schematic diagram of a process for generating an environmental risk prediction coefficient in an environmental dynamic monitoring system combining ecological meteorology and satellite remote sensing provided in an embodiment of the present application.

[0013] Explanation of the accompanying drawings: ecological meteorological monitoring module 10, satellite remote sensing monitoring module 20, data fusion module 30, risk prediction computing power analysis module 40, computing power driven risk prediction module 50, monitoring configuration adjustment module 60. DETAILED DESCRIPTION

[0014] The embodiment of the present application provides an environmental dynamic monitoring system that combines ecological meteorology with satellite remote sensing. The ecological meteorology monitoring module 10 and the satellite remote sensing monitoring module 20 are used to collect data, the data is integrated through the data fusion module 30, and then the risk prediction is performed through the risk prediction computing power analysis module 40 and the computing power driven risk prediction module 50. Finally, the monitoring configuration is dynamically adjusted according to the risk situation through the monitoring configuration adjustment module 60. The embodiment of the present application solves the technical problem that the existing environmental monitoring method has a single data source, resulting in insufficient environmental monitoring accuracy, and achieves the technical effect of improving the accuracy and reliability of environmental dynamic monitoring and risk prediction, as well as the rationality of monitoring resource allocation.

[0015] like Figure 1 As shown, the embodiment of the present application provides an environmental dynamic monitoring system combining ecological meteorology and satellite remote sensing, and the system includes:

[0016] The ecological meteorological monitoring module 10 is used to perform real-time monitoring of the target area and obtain an ecological meteorological monitoring data set.

[0017] Specifically, the target area is a specific geographical area where dynamic environmental monitoring is required. It can be a small ecological protection area, such as a forest reserve; it can also be a larger urban area or a river basin.

[0018] The ecological meteorological monitoring module 10 collects meteorological information related to the ecosystem in real time, including data such as temperature, humidity, wind speed, and precipitation, by arranging meteorological sensors such as thermometers, rain gauges, wind speed sensors, and humidity sensors in the target area. For example, temperature, humidity, and wind speed sensors are set in the forest area to record climate conditions at regular intervals. These data can be transmitted to the central data center via a wireless network and compiled into an ecological meteorological monitoring data set. The ecological meteorological monitoring data set reflects the real-time climate conditions and ecological environment status of the target area, which helps to understand the ecological characteristics of the area and its changing trends.

[0019] The satellite remote sensing monitoring module 20 is used to obtain a satellite remote sensing monitoring data set of the target area.

[0020] Specifically, the satellite remote sensing monitoring module 20 periodically uses sensors carried on the satellite to obtain data such as vegetation coverage, water changes, and land use in the target area. For example, optical sensors can obtain visible light and near-infrared band images of the target area to identify vegetation coverage, land use type, etc.; microwave sensors can penetrate clouds to monitor information such as the moisture content of the surface. Remote sensing data is transmitted to the ground station via satellite to form a satellite remote sensing monitoring data set, providing a wide range of environmental information in the target area.

[0021] The data fusion module 30 is used to fuse the ecological meteorological monitoring data set and the satellite remote sensing monitoring data set to generate an environmental dynamic monitoring result.

[0022] Specifically, the data fusion module 30 receives the ecological meteorological monitoring data set and the satellite remote sensing monitoring data set, and uses the data fusion algorithm to match and integrate the data from the two different sources in space and time, so as to generate the environmental dynamic monitoring results. The data fusion process usually includes data preprocessing, spatiotemporal matching, feature extraction and matching, and the final fusion output. Meteorological data is usually point data, including timestamps and coordinates, while satellite data is image data, which represents the state of the entire area. First, the ecological meteorology such as temperature, humidity, and wind speed is standardized, the dimension is removed, and the missing values ​​are filled; the remote sensing data is imaged and denoised. Then, the data is aligned in time and space, and the meteorological point data is matched with the data at the corresponding position and time in the remote sensing image to ensure that the two data correspond to each other at the same time and in the same area. Next, key features (such as high temperature, low humidity, etc.) are extracted from the ecological meteorological data, and relevant environmental features (such as vegetation index NDVI, water area, etc.) are extracted from the satellite remote sensing image. The meteorological features and remote sensing features are associated through feature matching to more comprehensively describe the regional environmental status. The integrated environmental dynamic monitoring results combine the detailed ecological and meteorological information on the ground and the macro information from satellite remote sensing, more comprehensively reflecting the environmental dynamic conditions of the target area, improving the comprehensiveness of environmental monitoring, and providing data support for subsequent risk prediction.

[0023] The risk prediction computing power analysis module 40 is used to perform risk prediction computing power analysis on the target area based on the environmental abnormality depth threshold and according to the environmental dynamic monitoring results to determine the environmental risk prediction computing power.

[0024] Specifically, the environmental anomaly depth threshold is a threshold used to identify the environmental state and to distinguish between normal and abnormal environmental data. The environmental risk prediction computing power indicates the scale of computing power required for risk prediction calculations. The greater the computing power, the more computing resources are needed to complete the environmental risk prediction task.

[0025] The risk prediction computing power analysis module 40 performs in-depth analysis on the environmental dynamic monitoring results to identify monitoring data that exceeds the environmental abnormality depth threshold, such as temperature, humidity or vegetation index that exceeds the normal range. According to the abnormality of the identified monitoring data, the computing power required for environmental risk prediction is determined, that is, the environmental risk prediction computing power. This computing power value will be used in the risk prediction model to generate a more accurate environmental risk assessment under a specific computing power.

[0026] The computing power-driven risk prediction module 50 is used to drive the environmental risk prediction channel to perform environmental risk prediction on the target area based on the environmental risk prediction computing power to obtain an environmental risk prediction coefficient.

[0027] Specifically, the environmental risk prediction channel is a path dedicated to risk prediction, which includes multiple different environmental risk prediction models, such as neural network models, support vector machine (SVM) models, etc. The environmental risk prediction channel can derive the environmental risk prediction coefficient of the target area based on the input environmental dynamic monitoring results. The environmental risk prediction coefficient is a numerical indicator used to indicate the size of the environmental risk faced by the target area. The larger the environmental risk prediction coefficient value, the higher the environmental risk of the target area.

[0028] The computing power-driven risk prediction module 50 receives the environmental risk prediction computing power from the risk prediction computing power analysis module 40. This computing power value represents the computing resources required for the current regional risk prediction task. According to the allocated computing power, the computing power-driven risk prediction module 50 will activate the corresponding environmental risk prediction channel, which includes multiple activated environmental risk prediction models, and input the environmental dynamic monitoring results into multiple activated environmental risk prediction models. Each model will derive an environmental risk coefficient based on its own algorithm and parameters. Combining these predicted environmental risk coefficients, a corresponding environmental risk prediction coefficient is generated. This environmental risk prediction coefficient can intuitively reflect the degree of environmental risk in the target area and provide a basis for subsequent environmental governance decisions.

[0029] The monitoring configuration adjustment module 60 is used to optimize the monitoring configuration of the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20 based on the first environmental risk threshold and the second environmental risk threshold of the target area and according to the environmental risk prediction coefficient, wherein the first environmental risk threshold is greater than the second environmental risk threshold.

[0030] Specifically, the first environmental risk threshold and the second environmental risk threshold are important indicators for determining the severity of environmental risks. The first environmental risk threshold represents a higher risk level, while the second environmental risk threshold represents a relatively lower risk level. Typically, the first threshold is higher than the second threshold.

[0031] The monitoring configuration adjustment module 60 receives the environmental risk prediction coefficient from the computing power driven risk prediction module 50, and compares it with the first environmental risk threshold and the second environmental risk threshold. If the environmental risk prediction coefficient is greater than or equal to the first threshold, it means that the risk faced by the area is high, and the monitoring frequency needs to be increased or the monitoring accuracy needs to be improved; if the environmental risk prediction coefficient is lower than the second threshold, it means that the risk is low, and the monitoring frequency can be reduced or the monitoring accuracy can be reduced; if the environmental risk prediction coefficient is between the two thresholds, a medium-intensity monitoring configuration is adopted. According to the threshold judgment result, the monitoring configuration adjustment module 60 will generate corresponding optimization instructions to adjust the monitoring parameters, monitoring frequency, monitoring range, etc. of the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20. These optimization instructions include increasing or decreasing the working intensity of the monitoring equipment, adjusting the monitoring area range, or changing the time interval for data collection. For example, in high-risk situations, it may be necessary to increase the frequency of obtaining satellite remote sensing images or increase the number of meteorological sensors; in low-risk situations, the monitoring frequency may be reduced accordingly.

[0032] By comparing and judging the environmental risk thresholds, the module can flexibly adjust the configuration parameters of the ecological meteorological monitoring module and the satellite remote sensing monitoring module 20 according to the different degrees of environmental risks, provide customized monitoring configurations for different risk levels, and improve resource utilization.

[0033] Further, such as Figure 2 As shown, the risk prediction computing power analysis module 40 described in the embodiment of the present application is also used to perform the following steps:

[0034] Step P41: Performing an environmental anomaly depth analysis on the target area according to the environmental dynamic monitoring result to obtain an environmental anomaly depth coefficient.

[0035] Step P42: Determine whether the environmental abnormality depth coefficient is greater than or equal to the environmental abnormality depth threshold.

[0036] Step P43: If the environmental anomaly depth coefficient is greater than or equal to the environmental anomaly depth threshold, activate the environmental risk prediction computing power analysis table in the risk prediction computing power analysis module 40, wherein the environmental risk prediction computing power analysis table includes multiple predetermined environmental anomaly depth intervals, and multiple predetermined environmental risk prediction computing powers corresponding to the multiple predetermined environmental anomaly depth intervals.

[0037] Step P44: Input the environmental anomaly depth coefficient into the environmental risk prediction computing power analysis table to obtain the environmental risk prediction computing power.

[0038] Specifically, the risk prediction computing power analysis module 40 uses a machine learning algorithm (such as a neural network or a decision tree) to analyze the results of dynamic environmental monitoring, analyze the environmental anomalies in the target area, and generate an environmental anomaly depth coefficient. The environmental anomaly depth coefficient is a quantitative indicator used to indicate the severity of the environmental anomaly in the target area. The larger the value, the more serious the environmental anomaly, which may be an extreme deviation in meteorological conditions or a significant imbalance in the ecosystem. For example, the target area is a forest. Recent meteorological data show that the temperature has risen abnormally, and satellite remote sensing data show that the vegetation coverage has dropped sharply. The change amplitude and change rate of these data are calculated through a comprehensive evaluation algorithm to obtain a value indicating the severity of the forest environmental anomaly, that is, the environmental anomaly depth coefficient.

[0039] The obtained environmental anomaly depth coefficient is compared with the pre-set environmental anomaly depth threshold. The environmental anomaly depth threshold is a criterion used to determine whether the environmental anomaly depth coefficient reaches the conditions for triggering further calculation and risk analysis, and can be set based on historical data, expert experience or relevant standards.

[0040] When the environmental anomaly depth coefficient is greater than or equal to the environmental anomaly depth threshold, it means that the environmental anomaly in the target area is more serious and a more in-depth risk prediction computing power analysis is required. At this time, the environmental risk prediction computing power parsing table is activated. This parsing table is a pre-built table that contains multiple predetermined environmental anomaly depth intervals and multiple predetermined environmental risk prediction computing powers corresponding to these intervals. Enter the environmental anomaly depth coefficient into the parsing table, and extract the corresponding environmental risk prediction computing power according to the depth interval to which the coefficient belongs. Among them, the predetermined environmental risk prediction computing power is a positive integer, which represents the number of environmental risk prediction models in the environmental risk prediction channel that need to be called for environmental risk prediction, and its value range is 1 to P, where P is the total number of models in the environmental risk prediction channel.

[0041] This module analyzes the environmental anomaly depth, obtains the environmental anomaly depth coefficient, and dynamically allocates computing power based on the environmental risk prediction computing power analysis table. It realizes flexible allocation of computing resources at different anomaly depths, makes environmental risk predictions more accurate and faster, and avoids unnecessary waste of computing power.

[0042] Furthermore, the risk prediction computing power analysis module 40 described in the embodiment of the present application is also used to perform the following steps:

[0043] Step P411: Load the environmental anomaly identification record set of the target area.

[0044] Step P412: Based on a support vector machine, the environmental anomaly recognition record set is trained and tested to construct an environmental anomaly recognition model.

[0045] Step P413: Input the environmental dynamic monitoring result into the environmental anomaly recognition model to obtain the environmental anomaly recognition result.

[0046] Step P414: Perform environmental anomaly depth evaluation based on the environmental anomaly recognition result to generate the environmental anomaly depth coefficient.

[0047] Specifically, the environmental anomaly recognition record set of the target area is loaded from the database. The environmental anomaly recognition record set is a data set used to train and verify the environmental anomaly recognition model, which contains historical environmental monitoring data and annotated anomalies. For example, ecological meteorological data (temperature, precipitation, wind speed, etc.) at different time points, satellite remote sensing data (land use changes, vegetation cover changes, etc.), and whether specific environmental anomalies occurred at that time (such as drought, floods, forest fires, etc.).

[0048] The support vector machine is trained and tested using the loaded environmental anomaly recognition record set to build an environmental anomaly recognition model. The model is used to identify and judge whether the environmental dynamic monitoring results are abnormal. In the training phase, a portion of the data in the record set (such as 70% of the data) is input into the support vector machine to allow the algorithm to learn the patterns in the data, such as learning what combination of meteorological and remote sensing data is associated with a specific environmental anomaly. Then in the testing phase, the remaining data (such as 30% of the data) is used to evaluate the accuracy of the model. By continuously adjusting the parameters of the model (such as kernel function, penalty factor, etc.), a model that can accurately identify environmental anomalies is constructed. For example, for identifying water pollution risks, the model can learn which water quality indicators (such as chemical oxygen demand, ammonia nitrogen content, etc.) and satellite remote sensing image features (such as water color, transparency, etc.) are related to pollution events.

[0049] After the model training is completed, the module inputs the real-time environmental dynamic monitoring data into the environmental anomaly recognition model. The model analyzes the input data according to the previously learned patterns to obtain the environmental anomaly recognition results. This result may be a classification result (such as whether there is an environmental anomaly, what type of environmental anomaly it is, for example, whether it is caused by meteorological disasters or human pollution), or it may be a probability value, indicating the possibility of a certain environmental anomaly.

[0050] After obtaining the environmental anomaly recognition result, it is necessary to further quantify the depth of the environmental anomaly. If the environmental anomaly recognition result is a classification result, it is necessary to quantify it according to the severity represented by different categories. For example, environmental anomalies are divided into three categories: mild, moderate, and severe, corresponding to different numerical ranges. If the environmental anomaly recognition result is a probability value, the environmental anomaly depth coefficient can be determined based on the probability. This coefficient is a value that comprehensively reflects the severity of the environmental anomaly, and can be used as a basis for subsequent judgments on whether to initiate more in-depth risk prediction computing power analysis and other operations.

[0051] By constructing an environmental anomaly recognition model based on support vector machines, this module can effectively use historical data to analyze the current environmental dynamic monitoring results, thereby obtaining accurate environmental anomaly recognition results and environmental anomaly depth coefficients, providing a key basis for matching environmental risk prediction computing power and helping to conduct more accurate environmental risk assessments.

[0052] Further, such as Figure 3 As shown, the computing power driven risk prediction module 50 described in the embodiment of the present application is also used to perform the following steps:

[0053] Step P51: Based on the environmental risk prediction computing power, feature activation is performed on the environmental risk prediction channel to obtain multiple activated environmental risk prediction models.

[0054] Step P52: Input the environmental dynamic monitoring results into the multiple activated environmental risk prediction models to obtain multiple predicted environmental risk coefficients.

[0055] Step P53: Calculate the centralized value of the multiple predicted environmental risk coefficients to generate the environmental risk prediction coefficient.

[0056] Specifically, first, based on the environmental risk prediction computing power generated previously, the module activates the corresponding number of environmental risk prediction models in the environmental risk prediction channel. After each model is activated, the real-time environmental dynamic monitoring results are input into these activated models, and each model analyzes the data separately to generate corresponding multiple predicted environmental risk coefficients. The centralized value of all predicted environmental risk coefficients is calculated to obtain a comprehensive environmental risk prediction coefficient. The centralized value can be calculated using a variety of methods, such as average, median or weighted average. For example, the average value is used to calculate, add all predicted environmental risk coefficients, and then divide them by the number of predicted environmental risk coefficients. The coefficients output by the three activated models are 5, 7, and 6, respectively. The module can calculate the average value of 6 as the environmental risk prediction coefficient. This environmental risk prediction coefficient is obtained by combining the prediction results of multiple models, which can more comprehensively and objectively reflect the environmental risk status of the target area and provide an important basis for subsequent environmental management decisions.

[0057] Furthermore, the steps of constructing the environmental risk prediction channel described in the embodiment of the present application include:

[0058] Step P5-1: Load the environmental risk record set of the target area.

[0059] Step P5-2: Perform similar region search based on the target region to obtain multiple sample regions.

[0060] Step P5-3: Conduct environmental risk record learning based on the multiple sample areas and establish multiple environmental risk predictors.

[0061] Step P5-4: Perform incremental learning on the multiple environmental risk predictors according to the environmental risk record set to generate multiple environmental risk prediction models.

[0062] Step P5-5: Connect the multiple environmental risk prediction models to generate the environmental risk prediction channel.

[0063] Specifically, the environmental risk record set is a collection of environmental data related to the target area in the past period of time and the corresponding environmental risk conditions. These data include ecological meteorological data (such as historical records of temperature, precipitation, wind speed, etc.), satellite remote sensing data (such as historical conditions of land use changes, vegetation cover changes, etc.), and environmental risk events associated with these data (such as the occurrence of droughts, floods, forest fires, etc.). Loading the environmental risk record set is the basis for building an environmental risk prediction channel, providing a data source for subsequent learning and model building.

[0064] Retrieve similar areas based on the environmental characteristics of the target area (such as climate conditions, topography, etc.). By comparing the environmental characteristics of a large number of areas, find multiple sample areas with similar environmental conditions to the target area. The judgment of similarity can be based on a variety of factors, such as geographical characteristics (topography, landforms, etc.), climate type, ecosystem type, etc. For example, if the target area is a mountain forest ecosystem, then similar areas may be other mountain forest areas with similar altitudes, vegetation types, and climate conditions. The environmental risk data of these sample areas can provide more reference and reference for the environmental risk prediction of the target area.

[0065] A machine learning algorithm is used to learn environmental risk records obtained from multiple sample areas (including various environmental data of the sample areas and the corresponding environmental risk results) to establish multiple environmental risk predictors. Each predictor focuses on a certain aspect of environmental risk or uses different algorithms for prediction. For example, a predictor may be based on a decision tree algorithm to specifically predict forest fire risks. The prediction model is established by learning the relationship between factors related to forest fires in the sample area (such as vegetation dryness, meteorological conditions, etc.) and whether a fire occurs. Another predictor may be based on a neural network algorithm to predict flood risks. The relationship between factors related to floods in the sample area (such as rainfall, terrain slope, etc.) and the occurrence of floods is learned.

[0066] Based on the multiple environmental risk predictors that have been established, these predictors are incrementally learned using the target area's own environmental risk record set. Incremental learning can make the predictor better adapt to the specific environmental characteristics of the target area. As the environmental risk records of the target area are continuously learned, the predictor gradually transforms into an environmental risk prediction model that is more in line with the actual situation of the target area. For example, the forest fire risk predictor initially established based on the sample area becomes a more accurate model for predicting forest fire risks in the target area after learning factors such as the specific vegetation type, local meteorological patterns, and human activities in the target area.

[0067] Connect multiple environmental risk prediction models generated through incremental learning to form a complete environmental risk prediction channel. This channel can integrate the prediction capabilities of multiple models and predict the environmental risks of the target area from different perspectives. For example, in a comprehensive environmental risk prediction channel, it includes risk prediction models for meteorological disasters (such as typhoons, rainstorms, etc.) and risk prediction models for ecological damage (such as deforestation, land degradation, etc.). By connecting these models, various environmental risks that the target area may face can be comprehensively evaluated.

[0068] By integrating the target area's own environmental risk record set and the experience of similar areas, establishing multiple environmental risk prediction models and connecting them, and constructing an environmental risk prediction channel, we can make full use of data from different sources and different prediction methods to improve the accuracy and reliability of environmental risk prediction.

[0069] Furthermore, environmental risk records are learned based on the multiple sample areas to establish multiple environmental risk predictors, including:

[0070] Step P5-31: Perform similarity evaluation on the multiple sample regions based on the target region, obtain multiple region similarity coefficients, and set multiple sample constraint capacities according to the multiple region similarity coefficients.

[0071] Step P5-32: Based on the multiple sample constraint capacities, perform environmental risk record retrieval on the multiple sample areas respectively to obtain multiple environmental risk record sample sets.

[0072] Step P5-33: Train and verify the multiple environmental risk record sample sets respectively to generate the multiple environmental risk predictors.

[0073] Specifically, the similarity of multiple sample areas is evaluated based on the environmental characteristics of the target area (such as precipitation, temperature, land use type, etc.). The similarity coefficient can be calculated by analyzing the historical environmental data of the target area and each sample area (such as climate data, precipitation data, ecological data, etc.). A common method is to use a distance metric (such as Euclidean distance or cosine similarity) to quantify the similarity between the target area and the sample area. Then, according to the calculated similarity coefficient, a sample constraint capacity is set for each sample area. The sample constraint capacity is the limit on the amount of data selected from each sample area for environmental risk record learning. For sample areas with high regional similarity coefficients, their data are more correlated with the target area, and a larger sample constraint capacity can be set to provide valuable information for environmental risk prediction in the target area. On the contrary, for sample areas with low regional similarity coefficients, a smaller sample constraint capacity is set to avoid introducing too much less relevant data. For example, if the regional similarity coefficient of a sample area is 0.8, its sample constraint capacity can be set to 800 environmental risk records; and if the regional similarity coefficient is 0.3, the sample constraint capacity is set to 300 environmental risk records.

[0074] Environmental risk records are retrieved for multiple sample areas respectively, and a corresponding number of records are selected from the environmental risk record database according to the sample constraint capacity set for each sample area to generate multiple environmental risk record sample sets. These environmental risk record sample sets are sample sets containing environmental risk information obtained through retrieval, providing training data for the environmental risk predictor. For example, for a sample area, the sample constraint capacity is set to 500 records, and the most relevant 500 records are selected from the environmental risk record database containing thousands of records. These records cover the ecological meteorological data, satellite remote sensing data and corresponding environmental risk events (such as natural disasters, ecological damage events, etc.) of the sample area, thereby forming an environmental risk record sample set.

[0075] For each sample set of environmental risk records, a suitable machine learning algorithm (such as decision tree, neural network, support vector machine, etc.) is used for training. A portion of the data in the sample set of environmental risk records (such as 70% of the data) is used as training data, and the remaining data (such as 30% of the data) is used as verification data. The training process uses the environmental data of the sample area and the actual risk records to let the model learn the rules of risk prediction and generate an environmental risk predictor. The model prediction effect is tested by verification data to ensure the accuracy and generalization ability of the model. For example, in a sample set of forest fire environmental risk records, the training data includes environmental data such as vegetation coverage, temperature, humidity, and whether a forest fire occurs. The algorithm continuously adjusts its own parameters (such as the node splitting rules of the decision tree, the weights of the neural network, etc.) to fit the relationship between these data as accurately as possible. If the model predicts the occurrence of forest fires with a high accuracy on the verification data, it means that the model has good generalization ability and can be used as an effective forest fire environmental risk predictor.

[0076] Through the above-mentioned training and validation process, an environmental risk predictor is generated for each sample set of environmental risk records. These predictors will predict the environmental risks of the target area from the perspective of different sample area data, providing a model basis for the subsequent accurate prediction of the environmental risks of the target area.

[0077] Furthermore, the monitoring configuration adjustment module 60 described in the embodiment of the present application is also used to perform the following steps:

[0078] Step P61: Collect the monitoring configuration parameters of the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20 respectively, and obtain ecological meteorological monitoring configuration information and satellite remote sensing monitoring configuration information.

[0079] Step P62: Determine whether the environmental risk prediction coefficient is greater than or equal to the first environmental risk threshold.

[0080] Step P63: If the environmental risk prediction coefficient is greater than or equal to the first environmental risk threshold, a monitoring configuration enhancement instruction is generated.

[0081] Step P64: Based on the monitoring configuration enhancement instruction, the ecological meteorological monitoring configuration information and the satellite remote sensing monitoring configuration information are enhanced to obtain an ecological meteorological monitoring configuration optimization plan and a satellite remote sensing monitoring configuration optimization plan.

[0082] Step P65: Based on the ecological meteorological monitoring configuration optimization scheme and the satellite remote sensing monitoring configuration optimization scheme, the monitoring configuration of the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20 are adjusted.

[0083] Specifically, relevant monitoring configuration parameters are obtained from the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20. For the ecological meteorological monitoring module 10, its monitoring configuration parameters include the distribution density of meteorological observation stations, the types of meteorological elements observed (such as whether to observe the details of wind speed, wind direction, temperature, humidity, air pressure, etc.), the time interval of observation, etc., which constitute the ecological meteorological monitoring configuration information. For the satellite remote sensing monitoring module 20, the monitoring configuration parameters may include the orbital parameters of the satellite (such as orbital altitude, inclination, etc.), the sensor type and its working parameters (such as resolution, band setting, etc.), the frequency of image acquisition, etc., which constitute the satellite remote sensing monitoring configuration information.

[0084] Compare the environmental risk prediction coefficient and the first environmental risk threshold to determine whether the severity of the current environmental risk has reached the level where the monitoring configuration needs to be adjusted. If the environmental risk prediction coefficient is greater than or equal to the first environmental risk threshold, it means that the target area faces a higher environmental risk and requires more detailed and more frequent monitoring. At this time, the monitoring configuration adjustment module 60 generates a monitoring configuration strengthening instruction. This instruction is a signal that the existing monitoring configuration needs to be improved.

[0085] For the enhancement of ecological meteorological monitoring configuration, the monitoring configuration enhancement instruction may specifically include: increasing the number of meteorological observation stations, expanding the geographical scope of observation, or shortening the observation time interval, etc. According to the monitoring configuration enhancement instruction, the ecological meteorological monitoring configuration information is enhanced to obtain the ecological meteorological monitoring configuration optimization plan. For example, the original observation of meteorological elements once an hour is changed to observation once every 30 minutes, and meteorological observation stations are added in some key areas.

[0086] For satellite remote sensing monitoring configuration enhancement, the monitoring configuration enhancement instruction may specifically include: adjusting the satellite's orbital parameters to increase the observation frequency of a specific area, replacing a higher resolution sensor, or increasing the sensor's operating band, etc. The satellite remote sensing monitoring configuration information is enhanced according to the monitoring configuration enhancement instruction, thereby obtaining a satellite remote sensing monitoring configuration optimization plan.

[0087] Finally, according to the obtained ecological meteorological monitoring configuration optimization scheme and satellite remote sensing monitoring configuration optimization scheme, the actual monitoring configuration adjustment is performed on the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20. For example, the layout and observation frequency of the meteorological observation station are adjusted according to the ecological meteorological monitoring configuration optimization scheme, and the relevant parameters of the satellite are adjusted according to the satellite remote sensing monitoring configuration optimization scheme, so as to improve the monitoring capability of the target area environment and provide more reliable monitoring data for environmental management, decision-making and risk response.

[0088] Furthermore, the monitoring configuration adjustment module 60 described in the embodiment of the present application is also used to perform the following steps:

[0089] Step P621: If the environmental risk prediction coefficient is less than the first environmental risk threshold, determine whether the environmental risk prediction coefficient is less than or equal to the second environmental risk threshold.

[0090] Step P622: If the environmental risk prediction coefficient is less than or equal to the environmental risk second threshold, a monitoring configuration weakening instruction is generated.

[0091] Step P623: Based on the monitoring configuration weakening instruction, the ecological meteorological monitoring configuration information and the satellite remote sensing monitoring configuration information are weakened to obtain the ecological meteorological monitoring configuration optimization plan and the satellite remote sensing monitoring configuration optimization plan.

[0092] Specifically, if the environmental risk prediction coefficient is less than the first environmental risk threshold, it is further determined whether the environmental risk prediction coefficient is less than or equal to the second environmental risk threshold. If the environmental risk prediction coefficient is less than or equal to the second environmental risk threshold, it means that the environmental risk of the current target area is low, and the monitoring intensity can be weakened to reduce resource consumption. In this case, the monitoring configuration adjustment module 60 will generate a monitoring configuration weakening instruction, including reducing the monitoring frequency, reducing the data sampling rate, reducing the monitoring range, or shutting down certain non-critical monitoring equipment. According to the generated weakening instruction, the configuration of the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20 is weakened accordingly, thereby obtaining an ecological meteorological monitoring configuration optimization scheme and a satellite remote sensing monitoring configuration optimization scheme. For example, the weakening instruction is to reduce the monitoring frequency, and the ecological meteorological monitoring module 10 adjusts the sampling frequency of temperature and humidity from once an hour to once every two hours. The satellite remote sensing monitoring module 20 reduces the frequency of regular imaging of a large area from once a day to once a week.

[0093] When the environmental risk is low, this module generates monitoring configuration weakening instructions and adjusts the existing monitoring configuration according to these instructions to optimize the input of monitoring resources and reduce the waste of resources.

[0094] In summary, the environmental dynamic monitoring system combining ecological meteorology and satellite remote sensing provided by the embodiment of the present application has the following technical effects:

[0095] The ecological meteorological monitoring module 10 provides real-time and detailed meteorological data of the target area, while the satellite remote sensing monitoring module 20 provides a wide range of macro-environmental information. The two types of data are integrated and processed by the data fusion module 30 to form an environmental dynamic monitoring result covering time and space multi-dimensional information, achieving all-round coverage of monitoring data and avoiding the lack of timeliness and spatial resolution of a single data source. The risk prediction computing power analysis module 40 automatically analyzes and classifies environmental anomalies through the environmental anomaly depth analysis model, and uses the computing power analysis table to convert the environmental anomaly depth coefficient into environmental risk prediction computing power. It can match the corresponding computing power resources according to the actual environmental conditions of the target area, and optimize the computing power resource utilization rate while ensuring the accuracy of environmental risk prediction. The computing power drives the risk prediction module 50, which activates the environmental risk prediction channel based on the environmental risk prediction computing power, obtains multiple activated environmental risk prediction models, and uses multiple activated environmental risk prediction models to predict environmental risks in the target area to obtain environmental risk prediction coefficients, which not only improves the accuracy and timeliness of environmental risk prediction, but also can adapt to different environmental risk conditions. The monitoring configuration adjustment module 60 optimizes the monitoring configuration of the ecological meteorological monitoring module 10 and the satellite remote sensing monitoring module 20 according to the first environmental risk threshold and the second environmental risk threshold of the target area and the environmental risk prediction coefficient, enhances the monitoring parameters for high-risk situations, and reduces the monitoring intensity in low-risk areas, ensuring that the monitoring system can adjust its monitoring strategy according to the current environmental risk level and achieve reasonable allocation of monitoring resources.

[0096] Overall, the embodiments of the present application achieve high-precision dynamic environmental monitoring and accurate risk assessment of the target area through the integration of ecological meteorological data and satellite remote sensing data, combined with computing power-driven risk prediction and dynamic configuration adjustment. It not only improves the accuracy and reliability of dynamic environmental monitoring and risk prediction, but also significantly enhances the monitoring system's ability to respond to environmental risks, improves the efficiency of monitoring resource utilization, and provides strong technical support for environmental protection, disaster warning and ecological management.

[0097] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. The environmental dynamic monitoring system combining ecological meteorology and satellite remote sensing is characterized by: The system comprises: An ecological meteorological monitoring module, which is used to monitor the target area in real time and obtain an ecological meteorological monitoring data set; A satellite remote sensing monitoring module, wherein the satellite remote sensing monitoring module is used to obtain a satellite remote sensing monitoring data set of the target area; A data fusion module, the data fusion module is used to fuse the ecological meteorological monitoring data set and the satellite remote sensing monitoring data set to generate an environmental dynamic monitoring result; A risk prediction computing power analysis module, wherein the risk prediction computing power analysis module is used to perform risk prediction computing power analysis on the target area based on an environmental anomaly depth threshold and according to the environmental dynamic monitoring result to determine the environmental risk prediction computing power; A computing power-driven risk prediction module, wherein the computing power-driven risk prediction module is used to drive an environmental risk prediction channel to perform environmental risk prediction on the target area based on the environmental risk prediction computing power to obtain an environmental risk prediction coefficient; A monitoring configuration adjustment module, wherein the monitoring configuration adjustment module is used to optimize the monitoring configuration of the ecological meteorological monitoring module and the satellite remote sensing monitoring module based on a first environmental risk threshold and a second environmental risk threshold of the target area and according to the environmental risk prediction coefficient, wherein the first environmental risk threshold is greater than the second environmental risk threshold.

2. The system according to claim 1, characterized in that The risk prediction computing power analysis module is used to perform risk prediction computing power analysis on the target area based on the environmental abnormality depth threshold and according to the environmental dynamic monitoring result to determine the environmental risk prediction computing power, including: Performing an environmental anomaly depth analysis on the target area according to the environmental dynamic monitoring result to obtain an environmental anomaly depth coefficient; Determining whether the environmental abnormality depth coefficient is greater than or equal to the environmental abnormality depth threshold; If the environmental anomaly depth coefficient is greater than or equal to the environmental anomaly depth threshold, the environmental risk prediction computing power analysis table in the risk prediction computing power analysis module is activated, wherein the environmental risk prediction computing power analysis table includes a plurality of predetermined environmental anomaly depth intervals, and a plurality of predetermined environmental risk prediction computing powers corresponding to the plurality of predetermined environmental anomaly depth intervals; The environmental anomaly depth coefficient is input into the environmental risk prediction computing power analysis table to obtain the environmental risk prediction computing power.

3. The system according to claim 2, characterized in that Performing an environmental anomaly depth analysis on the target area according to the environmental dynamic monitoring result to obtain an environmental anomaly depth coefficient includes: Loading the environmental anomaly identification record set of the target area; Based on a support vector machine, the environmental anomaly recognition record set is trained and tested to build an environmental anomaly recognition model; Inputting the environmental dynamic monitoring result into the environmental anomaly recognition model to obtain an environmental anomaly recognition result; An environmental anomaly depth evaluation is performed based on the environmental anomaly recognition result to generate the environmental anomaly depth coefficient.

4. The system according to claim 1, characterized in that The computing power driven risk prediction module is used to drive the environmental risk prediction channel to perform environmental risk prediction on the target area based on the environmental risk prediction computing power to obtain an environmental risk prediction coefficient, including: Based on the environmental risk prediction computing power, feature activation is performed on the environmental risk prediction channel to obtain multiple activated environmental risk prediction models; Inputting the environmental dynamic monitoring results into the multiple activated environmental risk prediction models to obtain multiple predicted environmental risk coefficients; Calculate the centralized value of the plurality of predicted environmental risk coefficients to generate the environmental risk prediction coefficient.

5. The system according to claim 1, wherein: The steps of constructing the environmental risk prediction channel include: Loading the environmental risk record set of the target area; Perform similar area search according to the target area to obtain multiple sample areas; Conducting environmental risk record learning based on the multiple sample areas to establish multiple environmental risk predictors; Performing incremental learning on the multiple environmental risk predictors according to the environmental risk record set to generate multiple environmental risk prediction models; The multiple environmental risk prediction models are connected to generate the environmental risk prediction channel.

6. The system according to claim 5, characterized in that Performing environmental risk record learning based on the multiple sample areas and establishing multiple environmental risk predictors include: Performing similarity evaluation on the plurality of sample regions based on the target region to obtain a plurality of region similarity coefficients, and setting a plurality of sample constraint capacities according to the plurality of region similarity coefficients; Based on the multiple sample constraint capacities, respectively search for environmental risk records of the multiple sample areas to obtain multiple environmental risk record sample sets; The multiple environmental risk record sample sets are trained and verified respectively to generate the multiple environmental risk predictors.

7. The system according to claim 1, characterized in that The monitoring configuration adjustment module is used to optimize the monitoring configuration of the ecological meteorological monitoring module and the satellite remote sensing monitoring module based on the first environmental risk threshold and the second environmental risk threshold of the target area and according to the environmental risk prediction coefficient, including: Collecting the monitoring configuration parameters of the ecological meteorological monitoring module and the satellite remote sensing monitoring module respectively, and obtaining ecological meteorological monitoring configuration information and satellite remote sensing monitoring configuration information; Determining whether the environmental risk prediction coefficient is greater than or equal to the environmental risk first threshold; If the environmental risk prediction coefficient is greater than or equal to the first environmental risk threshold, generating a monitoring configuration strengthening instruction; Based on the monitoring configuration enhancement instruction, the ecological meteorological monitoring configuration information and the satellite remote sensing monitoring configuration information are enhanced to obtain an ecological meteorological monitoring configuration optimization plan and a satellite remote sensing monitoring configuration optimization plan; Based on the ecological meteorological monitoring configuration optimization scheme and the satellite remote sensing monitoring configuration optimization scheme, the monitoring configuration of the ecological meteorological monitoring module and the satellite remote sensing monitoring module are adjusted.

8. The system according to claim 7, characterized in that Determining whether the environmental risk prediction coefficient is greater than or equal to the first environmental risk threshold includes: If the environmental risk prediction coefficient is less than the first environmental risk threshold, determining whether the environmental risk prediction coefficient is less than or equal to the second environmental risk threshold; If the environmental risk prediction coefficient is less than or equal to the environmental risk second threshold, generating a monitoring configuration weakening instruction; Based on the monitoring configuration weakening instruction, the ecological meteorological monitoring configuration information and the satellite remote sensing monitoring configuration information are weakened to obtain the ecological meteorological monitoring configuration optimization plan and the satellite remote sensing monitoring configuration optimization plan.

Images