Method and system for extracting plume region caused by heavy rainfall process through remote sensing

Abstract

The invention discloses a method and system for extracting a plume region caused by a heavy rainfall process through remote sensing, and belongs to the technical field of environmental science and monitoring. Identifying a heavy rainfall event; creating a plume region remote sensing index calculation formula, and calculating to obtain a plume region remote sensing extraction index PI caused by the heavy rainfall process of the current region; constructing a PI threshold inequality; determining a PI threshold range by using the PI threshold inequality and the water body remote sensing reflectivity; and applying the plume region remote sensing extraction index PI and the PI threshold range to other independent heavy rainfall events of the current region, and performing remote sensing extraction on the current region to obtain a corresponding plume region. In the construction process of the extraction index, by flexibly setting the threshold range, the method can adapt to different rainfall conditions and water body types, and limitation caused by a fixed suspended matter concentration range is avoided. The method not only can identify the water body in the turbid area, but also can effectively treat the turbid change of the water body under the condition of rainfall with different intensities.

Description

Technical Field

[0001] The present invention belongs to the technical field of environmental science and monitoring technology, and particularly relates to a method and system for remotely sensing and extracting a plume area caused by a heavy rainfall process. Background Art

[0002] With climate change and advancing urbanization, heavy rainfall events are occurring more frequently worldwide. Heavy rainfall can cause large amounts of surface water to flow into water bodies, particularly at lake estuaries, bringing with it large amounts of suspended matter, pollutants, and nutrients, creating areas of high turbidity (plumes). Plumes are typically characterized by a rapid increase in suspended matter concentrations and significant changes in water quality, which can have significant impacts on aquatic ecosystems and human activities, such as water quality monitoring and water resource utilization.

[0003] Traditional water monitoring methods, such as field sampling and laboratory analysis, often suffer from limited spatial coverage, long time periods, and high costs, making them difficult to meet the needs of large-scale, real-time monitoring. Remote sensing technology, with its ability to provide efficient and continuous monitoring data over large areas in a short period of time, has become a crucial tool in water environmental monitoring.

[0004] Methods for extracting turbidity zones in water bodies based on remote sensing data have made some progress. However, most existing technologies focus on monitoring static water bodies or over long timescales and have yet to effectively address the real-time monitoring of dynamic water changes (such as plume zones) caused by heavy rainfall. During heavy rainfall, suspended matter concentrations and water structure fluctuate significantly, especially over a short period of time. Consequently, there is a lack of high-precision, real-time extraction methods for this process.

[0005] In addition, existing methods usually rely on a fixed range of suspended matter concentration to identify plume areas. However, the suspended matter concentration in the plume area caused by heavy rainfall events varies greatly due to factors such as rainfall intensity, rainfall duration, and topography. This makes it difficult for fixed-range extraction methods to adapt to changes in rainfall conditions of different intensities. For example, under light rain and heavy rain, the changes in suspended matter concentration in water bodies vary significantly, and the diffusion and sedimentation rates of suspended matter in water bodies are also different. Therefore, how to accurately distinguish the plume area from the surrounding water body under large differences in rainfall intensity remains an important challenge in remote sensing water body monitoring. Summary of the Invention

[0006] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a method and system for remotely sensing the extraction of plume areas caused by heavy rainfall processes.

[0007] The present invention is achieved through the following technical solution: a method for remotely sensing and extracting plume areas caused by heavy rainfall processes, comprising the following steps: Collect valid satellite images of the corresponding weather station in the current region, obtain the cumulative rainfall in the N days before the valid satellite images, and identify heavy rainfall events based on the cumulative rainfall; Create a calculation formula for the plume area remote sensing index to calculate the plume area remote sensing extraction index PI caused by the current heavy rainfall process; A PI threshold inequality is constructed based on the heavy rainfall event that has been determined; and a PI threshold range is determined using the PI threshold inequality and the remote sensing reflectivity of the water body; Apply the plume area remote sensing extracted index PI and PI threshold range to other independent heavy rainfall events in the current region , remote sensing extraction of the current region to obtain the corresponding plume area .

[0008] In a further embodiment, the following steps are also included: Based on plume area Divide the current area into high turbidity areas and non-high turbidity areas, and build a verification model; The validation model is used to verify and optimize the effectiveness of the plume area remote sensing extraction index PI and PI threshold range.

[0009] In a further embodiment, the method for identifying the heavy rainfall event is as follows: The following formula is used to calculate the cumulative rainfall in the previous N days : Where, For the daily rainfall; like , it is determined to be a heavy rainfall event, and Q is the rainfall threshold.

[0010] In a further embodiment, the calculation formula of the plume area remote sensing index is expressed as follows: ; Where, 、 and Represent the wavelengths of the green light band, red light band and near-infrared band respectively, 、 and Green light band , red light band and near-infrared bands Remote sensing reflectivity.

[0011] In a further embodiment, the construction process of the PI threshold inequality is as follows: Select pixel remote sensing reflectivity data from the heavy rainfall events to distinguish the non-turbid area, the turbid area before the rainfall and the turbid area after the rainfall; The remote sensing reflectivity data of all image elements in the non-turbid area, the turbid area before rainfall and the turbid area after rainfall were extracted to obtain three sets of data sets, namely: clean water spectral data set , turbid water body spectral dataset before rainfall II and turbid water body spectral dataset after rainfall III; Draw box plots of the three data sets in different bands, and analyze the PI threshold inequality.

[0012] In a further embodiment, the PI threshold inequality is expressed as: Clean water spectral dataset : or Where, To clean the water spectral dataset The plume area remote sensing extraction index PI, 、 、 and To clean the water spectral dataset The mean remote sensing reflectance of the green band, the mean remote sensing reflectance of the near-infrared band, and the minimum and maximum remote sensing reflectance of the red band; Turbid water body spectral dataset before rainfall II: or Where, is the plume area remote sensing extracted index PI of the turbid water body pre-rainfall spectral dataset II, 、 、 and are the mean remote sensing reflectance of the green band, the mean remote sensing reflectance of the near-infrared band, and the minimum and maximum remote sensing reflectance of the red band of the spectral dataset II of turbid water bodies before rainfall; Turbid water body spectral dataset after rainfall III: and Where, is the remote sensing extracted index PI of the plume area of ​​the turbid water body after rainfall spectral dataset III, 、 、 and are the mean remote sensing reflectance of the green band, the mean remote sensing reflectance of the near-infrared band, and the minimum and maximum remote sensing reflectance of the red band of the spectral dataset III of turbid water bodies after rainfall; Correspondingly, the process of determining the PI threshold range is as follows: Use PI threshold inequality to obtain clean water spectral data sets , the clean water index range given by the turbid water body spectrum dataset II before rainfall and the turbid water body spectrum dataset III after rainfall , Turbid water body index range before rainfall and plume index range after rainfall in turbid water bodies ; The clean water index range , Turbid water body index range before rainfall The range of plume index after the clean water body is negated is obtained by negation The range of the plume index after the inversion of the turbid water body before rainfall and the inversion of the plume index , combined with the plume index range of the turbid water body after rainfall Take the intersection to get the comprehensive index range , get the comprehensive index range The minimum boundary value of and the maximum boundary value ; Then, the PI threshold range is: .

[0013] In a further embodiment, the plume region The extraction steps are as follows: Get other independent heavy rainfall events The effective observation image of the plume area remote sensing index calculation formula is used to obtain the heavy rainfall event Remote sensing extracted index of plume area ; Remote sensing extracted index of plume area If it falls within the PI threshold range, it is a plume area. The water body in the area corresponding to the effective observation image is the turbid water body, and the other areas are the non-turbid water bodies.

[0014] In a further embodiment, the construction process of the verification model is as follows: Based on the divided high turbidity area and non-high turbidity area, the fourth quantile value TSM of the turbidity area water body is calculated respectively. Q4 and the first quantile value TSM of water bodies in non-turbidity areas Q1 ; If TSM Q4 >TSM Q1 , it means that the plume area remote sensing extraction index PI is valid, otherwise it is invalid.

[0015] A system for remotely sensing and extracting plume areas caused by heavy rainfall processes, for implementing the above-mentioned method, comprises: The first module is configured to collect valid satellite images of a meteorological station corresponding to the current region, obtain the cumulative rainfall in the N days preceding the valid satellite images, and identify heavy rainfall events based on the cumulative rainfall; The second module is set to create a calculation formula for the plume area remote sensing index, and calculate the plume area remote sensing extraction index PI caused by the current regional heavy rainfall process; The third module is configured to construct a PI threshold inequality based on the heavy rainfall event determined; and determine the PI threshold range using the PI threshold inequality and the remote sensing reflectivity of the water body; The fourth module is configured to apply the plume area remote sensing extracted index PI and PI threshold range to other independent heavy rainfall events in the current region. , remote sensing extraction of the current region to obtain the corresponding plume area .

[0016] Beneficial effects of the present invention: (1) Flexible index threshold setting: This method can adapt to different rainfall conditions and water body types by flexibly setting the threshold range during the construction of the extraction index, avoiding the limitations of a fixed suspended matter concentration range. This method can not only identify turbid water bodies, but also effectively handle changes in water turbidity under different rainfall intensities.

[0017] (2) Reduce monitoring costs: Compared with traditional on-site sampling and laboratory analysis methods, the present invention can greatly reduce monitoring costs, reduce manual intervention and sampling workload, and achieve large-scale and high-frequency water quality monitoring through remote sensing data extraction.

[0018] (3) Broad application prospects: This method is applicable to the monitoring of various inland lakes and reservoirs and has good scalability. By processing remote sensing image data, it can provide real-time and accurate data support for water quality monitoring, water resource management, and ecological protection.

[0019] (4) Improve support for water management decision-making: Through efficient plume area extraction and water turbidity monitoring, the present invention provides strong technical support for water quality protection, pollution source tracking and ecological restoration, and can provide a scientific basis for water management departments to help make more accurate water environment management decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flow chart of the method for remote sensing extraction of plume areas caused by heavy rainfall processes in Example 1.

[0021] Figure 2 This is a schematic diagram of extracting three sets of data sets in Example 1.

[0022] Figure 3 It is a box plot of the corresponding bands of the three data sets in Example 1.

[0023] Figure 4 This is a comparison chart for validating the effectiveness of the plume area remote sensing extracted index PI in Example 1. DETAILED DESCRIPTION

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] Example 1 This embodiment takes Qiandao Lake as an example. The large surface runoff caused by heavy rainfall will carry a large amount of particulate matter, nutrients, pollutants, and heavy metals into the Xinanjiang Reservoir, thereby increasing the frequency and scale of algae outbreaks. Therefore, this embodiment discloses a method for remote sensing extraction of plume areas caused by heavy rainfall processes, such as Figure 1 As shown, the following steps are included: Collect valid satellite images of the corresponding weather station in the current region, obtain the cumulative rainfall in the N days before the valid satellite images, and identify heavy rainfall events based on the cumulative rainfall; Create a calculation formula for the plume area remote sensing index to calculate the plume area remote sensing extraction index PI caused by the current heavy rainfall process; A PI threshold inequality is constructed based on the heavy rainfall event that has been determined; and a PI threshold range is determined using the PI threshold inequality and the remote sensing reflectivity of the water body; Apply the plume area remote sensing extracted index PI and PI threshold range to other independent heavy rainfall events in the current region , remote sensing extraction of the current region to obtain the corresponding plume area .

[0026] In a further embodiment, the meteorological station corresponding to the current region can be the one closest to Qiandao Lake, and the valid satellite image is a remote sensing image acquired by the HJ-1A / B satellite in a cloud-free environment. The acquisition event of the remote sensing image is determined to be the current time, and the cumulative rainfall for the previous N days is obtained using the current time as the time point. In this embodiment, N is set to 7, meaning that the cumulative rainfall for the previous seven days is obtained (based on daily rainfall statistics).

[0027] Correspondingly, the identification method of heavy rainfall events is as follows: The following formula is used to calculate the cumulative rainfall in the previous N days : Where, For the daily rainfall; like , it is determined to be a heavy rainfall event, and Q is the rainfall threshold.

[0028] by 、 For example, if the cumulative rainfall in the first 7 days of the valid satellite image is , it is judged as a heavy rainfall event; otherwise, it is considered as a non-heavy rainfall event.

[0029] In a further embodiment, considering that the turbid water in the plume area will increase the remote sensing reflectivity of the red light band, the calculation formula of the plume area remote sensing index of this embodiment is expressed as follows: ; Where, 、 and Represent the wavelengths of the green light band, red light band and near-infrared band respectively, 、 and Green light band , red light band and near-infrared bands Remote sensing reflectivity.

[0030] Among them, the green light band is the green light of HJ-1A / B, The wavelength of the red light band is 660nm, and the wavelength of the near infrared band is 835nm. The calculation formula of the remote sensing index of the plume area is further expressed as: .

[0031] Based on this, the construction process of the PI threshold inequality is as follows: Pixel remote sensing reflectivity data were selected from those determined to be heavy rainfall events to distinguish between non-turbid areas, turbid areas before rainfall, and turbid areas after rainfall; the other 10 independent heavy rainfall events in Qiandao Lake were taken as examples.

[0032] Combine Figure 2 , the remote sensing reflectivity data of all image elements in the non-turbid area, turbid area before rainfall and turbid area after rainfall were extracted to obtain three sets of data sets, namely: clean water spectral data set , turbid water body spectral dataset before rainfall II and turbid water body spectral dataset after rainfall III; Each dataset contains the remote sensing reflectance of all pixels in the corresponding area in the green band, red band, and near-infrared band. The remote sensing reflectance data of all pixels in the non-turbid area, turbid area before rainfall, and turbid area after rainfall in the visible light and near-infrared bands are statistically analyzed, and box plots of the three datasets in different bands are drawn as follows: Figure 3 As shown, the PI threshold inequality is obtained analytically.

[0033] Based on the PI threshold inequality, the PI threshold range is determined as follows: Use PI threshold inequality to obtain clean water spectral data sets , the clean water index range given by the turbid water body spectrum dataset II before rainfall and the turbid water body spectrum dataset III after rainfall , Turbid water body index range before rainfall and plume index range after rainfall in turbid water bodies ; The clean water index range , Turbid water body index range before rainfall The range of plume index after the clean water body is negated is obtained by negation The range of the plume index after the inversion of the turbid water body before rainfall and the inversion of the plume index , combined with the plume index range of the turbid water body after rainfall Take the intersection to get the comprehensive index range , get the comprehensive index range The minimum boundary value of and the maximum boundary value ; Then, the PI threshold range is: In this embodiment, the comprehensive index range is The way to obtain is as follows: .

[0034] To achieve the above process, this embodiment further elaborates on the PI threshold inequality. First, Figure 3 (A) is the clean water spectral dataset The box plot of Figure 3 (B) is a box plot of the spectral dataset II of turbid water bodies before rainfall. Figure 3 (C) is a box plot of the spectral dataset III of turbid water bodies after rainfall.

[0035] Furthermore, for the clean water spectral dataset and the turbid water body pre-rainfall spectral dataset II, PI can be geometrically expressed as Figure 3 (A) and Figure 3 The negative value of the line segment 'BO' in (B); For the box plot of the spectral dataset III after rainfall in turbid water, PI can be geometrically expressed as Figure 3 The middle line segment 'BO' in (C) is shown in Figure 1. Based on the spectral distribution range of turbid water pixels in different datasets, the PI threshold inequality is constructed as follows: Clean water spectral dataset : or Where, To clean the water spectral dataset The plume area remote sensing extraction index PI, 、 、 and To clean the water spectral dataset The average remote sensing reflectance of the green light band, the average remote sensing reflectance of the near infrared band, and the minimum and maximum remote sensing reflectance of the red light band. The above scheme can determine the clean water index range in this embodiment. The plume index range after the corresponding clean water body is inverted , the inverse of which is the clean water index range other intervals.

[0036] Turbid water body spectral dataset before rainfall II: or Where, is the plume area remote sensing extracted index PI of the turbid water body pre-rainfall spectral dataset II, 、 、 and The turbid water body before rainfall index range in this embodiment can be determined by the above scheme. The corresponding inverted plume index range of turbid water body before rainfall .

[0037] Turbid water body spectral dataset after rainfall III: and Where, is the remote sensing extracted index PI of the plume area of ​​the turbid water body after rainfall spectral dataset III, 、 、 and The remote sensing reflectance of the green band, the remote sensing reflectance of the near-infrared band, and the minimum and maximum values ​​of the remote sensing reflectance of the red band of the turbid water body after rainfall spectral dataset III can be determined by the above scheme. .

[0038] Combined with the above description, the three interval ranges are intersected to obtain the minimum boundary value , the maximum boundary Furthermore, the PI threshold range in this embodiment is specifically expressed as follows: .

[0039] Based on the remote sensing extraction index PI and PI threshold range of the plume area, the plume area The extraction steps are as follows: Get other independent heavy rainfall events The effective observation image of the plume area remote sensing index calculation formula is used to obtain the heavy rainfall event Remote sensing extracted index of plume area ; Remote sensing extracted index of plume area If it falls within the PI threshold range, it is a plume area. The water body in the area corresponding to the effective observation image is the turbid water body, and the other areas are the non-turbid water bodies.

[0040] In another embodiment, the method further comprises: Divide the current area into high turbidity areas and non-high turbidity areas, and build a verification model; The validation model is used to verify and optimize the effectiveness of the plume area remote sensing extraction index PI and PI threshold range.

[0041] Furthermore, the construction process of the verification model is as follows: Based on the divided high turbidity area and non-high turbidity area, the fourth quantile value TSM of the turbidity area water body is calculated respectively. Q4 and the first quantile value TSM of water bodies in non-turbidity areas Q1 ; If TSM Q4 >TSM Q1 , it means that the plume area remote sensing extraction index PI is valid, otherwise it is invalid.

[0042] Combine Figure 4 Further explanation, the non-turbidity zone water body and the turbidity zone water body are distinguished by the measured total suspended matter concentration. According to the extraction results, the total suspended matter concentration of the turbidity zone water body (5.63±1.51 mg / L) is significantly higher than that of the non-turbidity zone water body (2.34±0.73 mg / L). In addition, in four independent heavy rainfall events, the first quantile value TSM of the non-turbidity zone water body is 1.37±1.70 mg / L. Q1 All are less than the fourth quantile value TSM of turbidity zone water body Q4 , indicating that the PI threshold (-0.062 <PI<0.016)能有效区分浑浊区水体和非浑浊区水体,该遥感提取指数有效。

[0043] Furthermore, for the results of verification being invalid, this embodiment also includes the following optimization method: Amplify the heavy rainfall events that have been identified, and obtain the corresponding valid satellite images to obtain the amplified sample data; for example, the corresponding amplified clean water spectral dataset , the number of pixels in the turbid water body pre-rainfall spectral dataset II and the turbid water body post-rainfall spectral dataset III. Quality control was also performed on the expanded spectral dataset to remove abnormal pixels such as clouds, shadows, and shoreline interference to ensure the representativeness and stability of the reflectance data.

[0044] Reconstruct the PI threshold inequality based on the amplified sample data, and use the reconstructed PI threshold inequality to determine the PI threshold range to obtain a new PI threshold range, that is, update the comprehensive index interval range The minimum boundary value of , comprehensive index range The maximum boundary value .

[0045] The new PI threshold range was used to extract other independent heavy rainfall events and perform TSM. Q4 >TSM Q1 Verification: If the verification passes, the new PI threshold range setting is retained; if it still fails, the sample data is repeatedly updated until a stable and effective PI threshold range is formed.

[0046] Example 2 This embodiment discloses a system for remotely sensing and extracting plume areas caused by heavy rainfall, which is used to implement the method described in Example 1, including: The first module is configured to collect valid satellite images of a meteorological station corresponding to the current region, obtain the cumulative rainfall in the N days preceding the valid satellite images, and identify heavy rainfall events based on the cumulative rainfall; The second module is set to create a calculation formula for the plume area remote sensing index, and calculate the plume area remote sensing extraction index PI caused by the current regional heavy rainfall process; The third module is configured to construct a PI threshold inequality based on the heavy rainfall event determined; and determine the PI threshold range using the PI threshold inequality and the remote sensing reflectivity of the water body; The fourth module is configured to apply the plume area remote sensing extracted index PI and PI threshold range to other independent heavy rainfall events in the current region. , remote sensing extraction of the current region to obtain the corresponding plume area ; The fifth module is set up based on the plume area Divide the current area into high turbidity areas and non-high turbidity areas, and build a verification model; The sixth module is configured to use the verification model to verify and optimize the effectiveness of the remote sensing extraction index PI and the PI threshold range of the plume area.

Claims

1. A method for remote sensing extraction of plume areas caused by heavy rainfall processes, characterized in that: The following steps are involved: Collect valid satellite images of the corresponding weather station in the current region, obtain the cumulative rainfall in the N days before the valid satellite images, and identify heavy rainfall events based on the cumulative rainfall; Create a calculation formula for the plume area remote sensing index to calculate the plume area remote sensing extraction index PI caused by the current heavy rainfall process; The PI threshold inequality is constructed based on the heavy rainfall events that have been determined; The PI threshold range is determined using the PI threshold inequality and the water body remote sensing reflectance; Apply the plume area remote sensing extracted index PI and PI threshold range to other independent heavy rainfall events in the current region , remote sensing extraction of the current region to obtain the corresponding plume area .

2. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 1, characterized in that: The following steps are also included: Based on plume area Divide the current area into high turbidity areas and non-high turbidity areas, and build a verification model; The validation model is used to verify and optimize the effectiveness of the plume area remote sensing extraction index PI and PI threshold range.

3. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 1, characterized in that: The method for identifying the heavy rainfall event is as follows: The following formula is used to calculate the cumulative rainfall in the previous N days : Where, For the daily rainfall; like , it is determined to be a heavy rainfall event, and Q is the rainfall threshold.

4. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 1, characterized in that: The calculation formula of the plume area remote sensing index is expressed as follows: ； Where, 、 and Represent the wavelengths of the green light band, red light band and near-infrared band respectively, 、 and Green light band , red light band and near-infrared bands Remote sensing reflectivity.

5. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 1, characterized in that: The construction process of the PI threshold inequality is as follows: Select pixel remote sensing reflectivity data from the heavy rainfall events to distinguish the non-turbid area, the turbid area before the rainfall and the turbid area after the rainfall; The remote sensing reflectivity data of all image elements in the non-turbid area, the turbid area before rainfall and the turbid area after rainfall were extracted to obtain three sets of data sets, namely: clean water spectral data set , turbid water body spectral dataset before rainfall II and turbid water body spectral dataset after rainfall III; Draw box plots of the three data sets in different bands, and analyze the PI threshold inequality.

6. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 5, characterized in that: The process of determining the PI threshold range is as follows: Use PI threshold inequality to obtain clean water spectral data sets , the clean water index range given by the turbid water body spectrum dataset II before rainfall and the turbid water body spectrum dataset III after rainfall , Turbid water body index range before rainfall and plume index range after rainfall in turbid water bodies ; The clean water index range , Turbid water body index range before rainfall The range of plume index after the clean water body is negated is obtained by negation The range of the plume index after the inversion of the turbid water body before rainfall and the inversion of the plume index , combined with the plume index range of the turbid water body after rainfall Take the intersection to get the comprehensive index range , get the comprehensive index range The minimum boundary value of and the maximum boundary value ; Then, the PI threshold range is: .

7. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 1, characterized in that: The plume region The extraction steps are as follows: Get other independent heavy rainfall events The effective observation images are used to calculate the heavy rainfall event using the remote sensing index calculation formula of the plume area. Remote sensing extracted index of plume area ; Remote sensing extracted index of plume area If it falls within the PI threshold range, it is a plume area. The water body in the area corresponding to the effective observation image is the turbid water body, and the other areas are the non-turbid water bodies.

8. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 2, characterized in that: The construction process of the verification model is as follows: Based on the divided high turbidity area and non-high turbidity area, the fourth quantile value TSM of the turbidity area water body is calculated respectively. Q4 and the first quantile value TSM of water bodies in non-turbidity areas Q1 ; If TSM Q4 >TSM Q1 , it means that the plume area remote sensing extraction index PI is valid, otherwise it is invalid and needs further optimization.

9. The method for remote sensing extraction of plume areas caused by heavy rainfall according to claim 2, characterized in that: The following optimization methods are also included: Amplify the heavy rainfall events that have been determined, and obtain the corresponding valid satellite images to obtain the amplified sample data; Reconstructing the PI threshold inequality based on the amplified sample data, and determining the PI threshold range using the reconstructed PI threshold inequality to obtain a new PI threshold range; The new PI threshold range was used to extract other independent heavy rainfall events and perform TSM. Q4 >TSM Q1 Verification: If the verification passes, the new PI threshold range setting is retained; if it still fails, the sample data is repeatedly updated until a stable and effective PI threshold range is formed.

10. A system for remote sensing extraction of plume areas caused by heavy rainfall, used to implement the method according to any one of claims 1 to 9, characterized in that: include: The first module is configured to collect valid satellite images of a meteorological station corresponding to the current region, obtain the cumulative rainfall in the N days preceding the valid satellite images, and identify heavy rainfall events based on the cumulative rainfall; The second module is set to create a calculation formula for the plume area remote sensing index, and calculate the plume area remote sensing extraction index PI caused by the current regional heavy rainfall process; The third module is set to construct the PI threshold inequality based on the heavy rainfall events that have been determined; The PI threshold range is determined using the PI threshold inequality and the water body remote sensing reflectance; The fourth module is configured to apply the plume area remote sensing extracted index PI and PI threshold range to other independent heavy rainfall events in the current region. , remote sensing extraction of the current region to obtain the corresponding plume area .

Images