River runoff length retention rate calculation method based on remote sensing extraction of water body

Abstract

The invention discloses a river runoff length retention rate calculation method based on remote sensing extraction of a water body. The river runoff length retention rate calculation method comprises the following specific steps: S1, preliminarily selecting a reference year of an evaluation year; s2, comparing remote sensing images to determine a final reference year; s3, extracting runoff length through visual interpretation: querying multi-period remote sensing image data, and calculating runoff length in a current year and a river channel in an evaluation year through visual interpretation and a GIS plotting mode; s4, calculating a result, wherein the runoff length retention rate is equal to the percentage of the runoff length in the dry season of the river evaluation year to the maximum runoff length in the dry season of the reference year; the method has the advantages of being efficient, accurate, comprehensive, dynamic, sustainable and the like, and powerful data support and decision basis can be provided for drainage basin water resource management, river ecological protection, environment monitoring and the like; the popularization and application of the method are beneficial to improving the precision of hydrological research and optimizing the scheduling and distribution of water resources, thereby promoting more scientific and sustainable water resource management.

Description

Technical Field

[0001] This invention relates to the field of hydrological technology, and in particular to a method for calculating the retention rate of river runoff length based on remote sensing extraction of water bodies. Background Technology

[0002] For rivers and lakes with approved ecological water volume protection implementation plans, the approved results should be directly adopted for determining the ecological water volume target value. For those without approval, the "Specification for Calculating Water Demand for River and Lake Ecological Environment" (SL / Z712-2021) can be referenced, or the percentage of the minimum monthly average flow from June to October and November to May can be calculated as the percentage of the multi-year monthly average flow for the corresponding periods. In accordance with the requirements of the "Notice of the Department of Rivers and Lakes of the Ministry of Water Resources on Further Clarifying Relevant Matters Concerning River and Lake Health Assessment" (River and Lake

[2023] No. 1), for seasonal rivers where the ecological flow is not clearly defined, scores can be assigned based on the runoff length retention rate.

[0003] Runoff length retention rate refers to the percentage of the dry season runoff length in the evaluation year relative to the maximum dry season runoff length in the reference year. The dry season runoff length in the evaluation year refers to the maximum length of the river section with water during the dry season of the evaluation year. The reference year should preferably be a year with a similar hydrological frequency to the evaluation year, following the promulgation of the "Regulations on River Management of the People's Republic of China" in 1988. Runoff length can be evaluated using flood survey data, remote sensing interpretation, or data published by relevant departments. Flood survey methods refer to the "Hydrological Survey Specifications" (SL 196).

[0004] SAR remote sensing imagery is used to extract water body information. Compared to optical imagery, SAR images are unaffected by weather and are most frequently used in emergency situations. For water bodies, during floods, the weather is generally overcast and cloudy, resulting in poor image quality for optical images. Therefore, SAR imagery is primarily used to assess the affected area. When interpreting runoff length using remote sensing, the river surface area retention rate can also be calculated, which is the percentage of the maximum water surface area during the dry season of the assessment year to the maximum water surface area during the dry season of the reference year. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for calculating the retention rate of river runoff based on remote sensing extraction of water bodies.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: The method for calculating the retention rate of river runoff length based on remote sensing extraction of water bodies includes the following specific steps: S1: Preliminary selection of the reference year for the evaluation year: The reference year should be a year with a hydrological frequency similar to that of the evaluation year; S1a: City A: The evaluation year is 2023. When determining the reference year, since there is a lack of existing long-term average precipitation data for the City A region, monthly precipitation data with a global land surface resolution of 0.5° produced by the National Centre for Atmospheric Sciences (NCAS) of the United Kingdom were selected. Monthly precipitation data for a portion of the City A region from 1980 to 2023 were extracted. Using precipitation frequency matching software developed by Wuhan University, the long-term precipitation data was input to plot the precipitation frequency P3 curve. Based on the fitting results, two years with precipitation close to 741 mm in 2023 were selected: 670 mm in 2016 and 862 mm in 2021. S1b: City B: The evaluation year is 2023. When determining the reference year, the precipitation frequency matching software developed by Wuhan University is used to input long-series precipitation data to draw the precipitation frequency P3 curve. Based on the fitting results, years with precipitation close to that of 2023 are selected between 2018 and 2023. S2: Determining the final reference year through comparison of remote sensing images: Based on remote sensing platforms such as National Geographic Cloud and AIEARTH, the final reference year is selected from the years 2018-2023 of existing remote sensing image data. S2a: City A: 2021, with a similar annual precipitation frequency and available remote sensing imagery, was ultimately selected and evaluated as the reference year; S2b: City B: 2020, which was selected and evaluated as the reference year with similar annual precipitation frequency and available remote sensing image data, was selected as the reference year, and October, which had available image data, was selected as the dry season. S3: Visual interpretation to extract runoff length: Query multi-period remote sensing image data and calculate the current year and evaluation year runoff length in the river channel through visual interpretation and GIS plotting; S4: Calculation results: Runoff length retention rate = percentage of the dry season runoff length in the river evaluation year to the maximum runoff length in the dry season of the reference year; S4a: City A: The current runoff length of City A in the dry season of 2023 was calculated to be 43.75 km using the above method. Compared with the runoff length of 44.19 km in the dry season of 2021, the ratio is 99.01%, that is, the runoff length retention rate is 99.01%. The ecological water volume satisfaction level is replaced by the runoff retention rate length. The ecological flow satisfaction level score of the river section of runoff A is 100 points. S4b: City B: The current runoff length of City B in the dry season of 2023 was calculated to be 6.366 km using the above method. Compared with the runoff length of 7.207 km in the dry season of 2020, the calculated result is 88.33%, that is, the runoff length retention rate is 88.33%. Using the runoff retention rate length to replace the ecological water volume satisfaction, the ecological flow satisfaction score of the river section evaluated by runoff B is 95.83 points.

[0007] As a further technical solution of the present invention, there is no hydrological station set up on the runoff A of city A, there is no hydrological observation data for the river, and there is no approved ecological flow for the ecological water volume guarantee implementation plan. Therefore, it is a seasonal river with no clear ecological flow and can be scored according to the runoff length retention rate.

[0008] As a further technical solution of the present invention, there is currently no hydrological station on runoff B in city B, and no approved ecological flow for ecological water volume guarantee implementation plan. After reviewing relevant hydrological data, consulting local residents and water authorities, and conducting on-site surveys, it was found that the data did not meet the conditions for calculating the ecological flow satisfaction level. Therefore, it is a seasonal river with no clear ecological flow, and scores can be assigned based on the runoff length retention rate.

[0009] As a further technical solution of the present invention, in S2, the remote sensing image adopts Sentinel-1 GRD data.

[0010] As a further technical solution of the present invention, the preprocessing flow of Sentinel-1 GRD data is generally as follows: orbit correction, thermal noise removal, radiometric calibration, multi-view processing, coherent filtering, geocoding, and decibel conversion.

[0011] As a further technical solution of the present invention, after the Sentinel-1 GRD data has undergone the corresponding preprocessing, it can be used for subsequent research on the extraction of water body target information. The threshold method of radar image water body extraction is used to perform corresponding processing on the original band or the corresponding water body index.

[0012] The beneficial effects of this invention are: it has the advantages of being efficient, accurate, comprehensive, dynamic, and sustainable, and can provide strong data support and decision-making basis for watershed water resources management, river ecological protection, and environmental monitoring; the promotion and use of this method will help improve the accuracy of hydrological research, optimize the scheduling and allocation of water resources, and thus promote more scientific and sustainable water resources management. Attached Figure Description

[0013] Figure 1 Precipitation map for City A from 1981 to 2023; Figure 2 This is a graph showing the fitting process of precipitation series data and precipitation frequency curve for City A. Figure 3 The process of screening and extracting remote sensing images of City A Figure 1 ; Figure 4 The process of screening and extracting remote sensing images of City A Figure 2 ; Figure 5 This is a curve showing the frequency of runoff and rainfall in area B of city B. Figure 6 Remote sensing image of the section of river B with water flow in City B in October 2023 (urban section / suburban section); Figure 7 Remote sensing image of the section of river B with water flow in City B in October 2020 (urban section / suburban section); Figure 8 This is a flowchart of a method for calculating the retention rate of river runoff length based on remote sensing extraction of water bodies. Detailed Implementation

[0014] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0015] Example 1 The runoff A of City A is the Shijin Main Canal in Shijiazhuang City. There is currently no hydrological station on the Shijin Main Canal in Shijiazhuang City, and there is no river with hydrological observation data. There is no ecological flow in the approved ecological water volume guarantee implementation plan. Therefore, it is a seasonal river with no clear ecological flow and can be scored according to the runoff length retention rate.

[0016] Please see Figure 1-4 Method 8, based on remote sensing extraction of water body water length retention rate, includes the following specific steps: S1: Preliminary selection of the reference year for the evaluation year: The reference year should be a year with a hydrological frequency similar to that of the evaluation year; S1a: City A: The evaluation year is 2023. When determining the reference year, since there is a lack of existing long-term average precipitation data for the City A region, monthly precipitation data with a global land surface resolution of 0.5° produced by the National Centre for Atmospheric Sciences (NCAS) of the United Kingdom were selected. Monthly precipitation data for a portion of the City A region from 1980 to 2023 were extracted. Using precipitation frequency matching software developed by Wuhan University, the long-term precipitation data was input to plot the precipitation frequency P3 curve. Based on the fitting results, two years with precipitation close to 741 mm in 2023 were selected: 670 mm in 2016 and 862 mm in 2021. S2: Determining the final reference year through comparison of remote sensing images: Based on remote sensing platforms such as National Geographic Cloud and AIEARTH, the final reference year is selected from the years 2018-2023 of existing remote sensing image data. S2a: City A: 2021, with a similar annual precipitation frequency and available remote sensing imagery, was ultimately selected and evaluated as the reference year; The remote sensing imagery uses Sentinel-1 GRD data. The Sentinel series of satellites is one of the Earth observation satellite series of the European Space Agency's "Copernicus Project". The Sentinel-1 data consists of two satellites, Sentinel-1A and Sentinel-1B, which carry C-band synthetic aperture radar. The replay period for the two satellites is 6 days.

[0017] The preprocessing workflow for Sentinel-1 GRD data generally includes: orbit correction, thermal noise removal, radiometric calibration, multi-view processing, coherent filtering, geocoding, and decibel conversion.

[0018] Orbit correction: Since the orbital status data in the default metadata file of Sentinel-1 raw imagery is not very accurate, this step requires an accurate orbital file; therefore, an accurate orbital file is downloaded from the Internet using SNAP to update the Sentinel-1 satellite orbital status information in real time.

[0019] Thermal noise removal: Thermal noise is the noise carried by the SAR satellite system itself. The distance that the SAR antenna travels from emitting electromagnetic waves to receiving electromagnetic waves is about 1400km. Due to the spherical diffusion effect of waves, the energy attenuates inversely proportional to the square of the distance. The transmitter needs huge power and emits extremely strong energy. Therefore, the heat loss inside the SAR satellite device cannot be ignored. In the SNAP software, the S-1 Thermal Noise module is selected to remove the thermal noise of the Sentinel-1 raw image.

[0020] Radiometric calibration: Radiometric calibration refers to converting the received backscattered signal into a physical quantity with units, where the physical quantity with units is the backscattering coefficient. Radiometric calibration is performed using SNAP software.

[0021] Multi-view processing: The entire effective synthetic aperture length is divided into 5 segments, each of which images the same scene. The resulting images are then summed and superimposed to obtain a SAR image, which improves the signal-to-noise ratio of the SAR image, suppresses speckle noise, and enhances image interpretability. GRD data has already undergone multi-view processing, and its pixels represent square ground areas. Although further multi-view processing can be performed because multi-view can eliminate or reduce the influence of speckle, multi-view will reduce the resolution of the image.

[0022] Coherent filtering: The commonly used coherent speckle filter is the Refined Lee filter (an improved Lee filter). It is an adaptive filter whose filtering window can be adjusted according to the region (usually 7 * 7, 11 * 11, etc.), and the processing effect is relatively good. The improved Lee filtering method can retain more edge information, which will slightly reduce the removal effect of speckle noise, and is more in line with the requirements of water body extraction for preserving edge details.

[0023] Geocoding: SAR images have three inherent geometric distortions caused by the difference between slant range and horizontal distance (ground distance). Two of these distortions are SAR-specific phenomena called "perspective shortening" and "top-bottom inversion," which are the main causes of geometric distortion. These two phenomena are particularly noticeable in mountainous areas. Topographic correction can effectively eliminate these phenomena.

[0024] Decibel conversion: The backscattering coefficient obtained after the above processing is a linear proportional unit, which is usually a small positive value. Due to the long transmission distance of the receiver, the radar backscattering received by the receiver is very small or the power is very small. The backscattering coefficient is logarithmically transformed, which is the decibel conversion.

[0025] After appropriate preprocessing, Sentinel-1 GRD data can be used for subsequent water target information extraction studies. The threshold method, a radar image water extraction method, is used to process the original band or the corresponding water index.

[0026] Sentinel-1 radar imagery also includes targeted water body indices, with the SDWI index being a commonly used one. The SDWI index enhances water body characteristics, making water bodies more clearly distinguishable while eliminating interference from soil and vegetation in water body extraction. The formula for the water body index involves multiplying the VV and VH polarimetric images and then multiplying by 10 to amplify the difference between water bodies and other land features, using the natural logarithm as the function.

[0027] After calculating the SDWI index, the Otsu method was used for threshold segmentation and subsequent evaluation of the accuracy of water body classification.

[0028] In mountainous areas, mountain shadows can easily be confused with water bodies. Further utilization of DEM data is necessary. Considering the slope characteristics of flood retention areas and the slope conditions of the mountain shadow areas formed by the Sentinel-1 image incident angle range of 29-46° in the study area, a slope threshold can be determined to exclude non-water areas (generally, a slope threshold of 10° is used; values ​​greater than 10° are removed, but this needs to be set according to the specific study area). This threshold can effectively mask and remove mountain shadows caused by SAR side-view imaging while preserving water information in flood retention areas.

[0029] S3: Visual interpretation to extract runoff length: Query multi-period remote sensing image data and calculate the current year and evaluation year runoff length in the river channel through visual interpretation and GIS plotting; S4: Calculation results: Runoff length retention rate = percentage of the dry season runoff length in the river evaluation year to the maximum runoff length in the dry season of the reference year; The scoring criteria for runoff length retention rate are shown in the table below.

[0030] Runoff Length Retention Rate Scoring Standard Table S4a: City A: The current runoff length of City A in the dry season of 2023 was calculated to be 43.75 km using the above method. Compared with the runoff length of 44.19 km in the dry season of 2021, the ratio is 99.01%, that is, the runoff length retention rate is 99.01%. The ecological water volume satisfaction level is replaced by the runoff retention rate length. The ecological flow satisfaction level score of the river section of runoff A is 100 points. Example 2 The runoff of City B is the Chengxi River in Wanquan District of Zhangjiakou City. There is currently no hydrological station on the Chengxi River in Wanquan District of Zhangjiakou City, and there is no approved ecological flow for the ecological water volume guarantee implementation plan. After reviewing relevant hydrological data, consulting local residents and water authorities, and conducting on-site surveys, it was found that the data did not meet the conditions for calculating the ecological flow satisfaction level. Therefore, it is a seasonal river with no clearly defined ecological flow, and scores can be assigned based on the runoff length retention rate.

[0031] Please see Figure 5-8 The method for calculating the retention rate of river runoff length based on remote sensing extraction of water bodies includes the following specific steps: S1: Preliminary selection of the reference year for the evaluation year: The reference year should be a year with a hydrological frequency similar to that of the evaluation year; S1b: City B: The evaluation year is 2023. When determining the reference year, the precipitation frequency matching software developed by Wuhan University is used to input long-series precipitation data to draw the precipitation frequency P3 curve. Based on the fitting results, years with precipitation close to that of 2023 are selected between 2018 and 2023. S2: Determining the final reference year through comparison of remote sensing images: Based on remote sensing platforms such as National Geographic Cloud and AIEARTH, the final reference year is selected from the years 2018-2023 of existing remote sensing image data. S2b: City B: 2020, which was selected and evaluated as the reference year with similar annual precipitation frequency and available remote sensing image data, was selected as the reference year, and October, which had available image data, was selected as the dry season. S3: Visual interpretation to extract runoff length: Query multi-period remote sensing image data and calculate the current year and evaluation year runoff length in the river channel through visual interpretation and GIS plotting; S4: Calculation results: Runoff length retention rate = percentage of the dry season runoff length in the river evaluation year to the maximum runoff length in the dry season of the reference year; S4b: City B: The current runoff length of City B in the dry season of 2023 was calculated to be 6.366 km using the above method. Compared with the runoff length of 7.207 km in the dry season of 2020, the calculated result is 88.33%, that is, the runoff length retention rate is 88.33%. Using the runoff retention rate length to replace the ecological water volume satisfaction, the ecological flow satisfaction score of the river section evaluated by runoff B is 95.83 points.

[0032] Results of the ecological flow satisfaction of runoff in City B As can be seen from the above description, the embodiments of the present invention achieve the following technical effects: they have the advantages of being efficient, accurate, comprehensive, dynamic, and sustainable, and can provide strong data support and decision-making basis for watershed water resources management, river ecological protection, and environmental monitoring; the promotion and use of this method will help improve the accuracy of hydrological research, optimize the scheduling and allocation of water resources, and thus promote more scientific and sustainable water resources management.

[0033] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples; within the framework of the invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of the different aspects of the invention as described above, which are not provided in the details for the sake of brevity.

[0034] This invention is intended to cover all such substitutions, modifications, and variations falling within the broad scope of the claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for calculating the retention rate of river runoff length based on remote sensing extraction of water bodies, characterized in that, The specific steps include the following: S1: Preliminary selection of the reference year for the evaluation year: The reference year should be a year with a hydrological frequency similar to that of the evaluation year; S1a: City A: The evaluation year is 2023. When determining the reference year, since there is a lack of existing long-term average precipitation data for the City A region, monthly precipitation data with a global land surface resolution of 0.5° produced by the National Centre for Atmospheric Sciences (NCAS) of the United Kingdom were selected. Monthly precipitation data for a portion of the City A region from 1980 to 2023 were extracted. Using precipitation frequency matching software developed by Wuhan University, the long-term precipitation data was input to plot the precipitation frequency P3 curve. Based on the fitting results, two years with precipitation close to 741 mm in 2023 were selected: 670 mm in 2016 and 862 mm in 2021. S1b: City B: The evaluation year is 2023. When determining the reference year, the precipitation frequency matching software developed by Wuhan University is used to input long-series precipitation data to draw the precipitation frequency P3 curve. Based on the fitting results, years with precipitation close to that of 2023 are selected between 2018 and 2023. S2: Determining the final reference year through comparison of remote sensing images: Based on remote sensing platforms such as National Geographic Cloud and AIEARTH, the final reference year is selected from the years 2018-2023 of existing remote sensing image data. S2a: City A: 2021, with a similar annual precipitation frequency and available remote sensing imagery, was ultimately selected and evaluated as the reference year; S2b: City B: 2020, which was selected and evaluated as the reference year with similar annual precipitation frequency and available remote sensing image data, was selected as the reference year, and October, which had available image data, was selected as the dry season. S3: Visual interpretation to extract runoff length: Query multi-period remote sensing image data and calculate the current year and evaluation year runoff length in the river channel through visual interpretation and GIS plotting; S4: Calculation results: Runoff length retention rate = percentage of the dry season runoff length in the river evaluation year to the maximum runoff length in the dry season of the reference year; S4a: City A: The current runoff length of City A in the dry season of 2023 was calculated to be 43.75 km using the above method. Compared with the runoff length of 44.19 km in the dry season of 2021, the ratio is 99.01%, that is, the runoff length retention rate is 99.01%. The ecological water volume satisfaction level is replaced by the runoff retention rate length. The ecological flow satisfaction level score of the river section of runoff A is 100 points. S4b: City B: The current runoff length of City B in the dry season of 2023 was calculated to be 6.366 km using the above method. Compared with the runoff length of 7.207 km in the dry season of 2020, the calculated result is 88.33%, that is, the runoff length retention rate is 88.33%. Using the runoff retention rate length to replace the ecological water volume satisfaction, the ecological flow satisfaction score of the river section evaluated by runoff B is 95.83 points.

2. The method for calculating the river runoff length retention rate based on remote sensing extraction of water bodies according to claim 1, characterized in that, There is currently no hydrological station on runoff A in City A, and there is no hydrological observation data for the river. There is no approved ecological flow for the ecological water volume guarantee implementation plan. Therefore, it is a seasonal river with no clear ecological flow and can be scored based on the runoff length retention rate.

3. The method for calculating the river runoff length retention rate based on remote sensing extraction of water bodies according to claim 1, characterized in that, There is currently no hydrological station on runoff B in City B, and there is no approved ecological flow for the ecological water volume guarantee implementation plan. After reviewing relevant hydrological data, consulting local residents and water authorities, and conducting on-site surveys, it was found that the data did not meet the conditions for calculating the ecological flow satisfaction level. Therefore, it is a seasonal river with no clearly defined ecological flow, and scores can be assigned based on the runoff length retention rate.

4. The method for calculating the river runoff length retention rate based on remote sensing extraction of water bodies according to claim 1, characterized in that, In S2, the remote sensing image uses Sentinel-1 GRD data.

5. The method for calculating the river runoff length retention rate based on remote sensing extraction of water bodies according to claim 4, characterized in that, The preprocessing workflow for Sentinel-1 GRD data generally includes: orbit correction, thermal noise removal, radiometric calibration, multi-view processing, coherent filtering, geocoding, and decibel conversion.

6. The method for calculating the river runoff length retention rate based on remote sensing extraction of water bodies according to claim 5, characterized in that, After appropriate preprocessing, Sentinel-1 GRD data can be used for subsequent water target information extraction studies. Thresholding methods for water extraction from radar images are used to process the original bands or corresponding water indices.

Images