Flood process runoff monitoring method coupled with satellite remote sensing and hydrodynamics

Abstract

The invention discloses a flood process runoff monitoring method coupled with satellite remote sensing and hydrodynamics. The flood process runoff monitoring method comprises the following steps that unmanned aerial vehicle complete river topography measurement is carried out; river upstream and downstream section hydraulic parameters are obtained; constructing a digital river channel section; batch extraction of satellite remote sensing water surface width; flow calculation parameters are obtained based on the digital river section and the water surface width data; performing remote sensing inversion on the roughness of the complete riverway; and section flow monitoring and historical data inversion. A section of complete river channel is determined as a basic research unit, flow calculation precision is remarkably improved through multi-section synchronous monitoring in combination with river channel slope and roughness change dynamically captured in real time, the flowing process of water flow in the complete river channel is finely described, and the flow calculation precision is improved. The method provides long-sequence and high-reliability flow data for parameter calibration and result verification of a hydrological model, is particularly suitable for describing the flow condition when flood comes, and enables the flow condition to be closer to reality.

Description

Technical Field

[0001] This invention relates to the fields of satellite remote sensing runoff monitoring and water conservancy engineering technology, particularly to river flood control and the acquisition of river runoff data in rivers without hydrological stations, specifically a method for monitoring flood process runoff that couples satellite remote sensing and hydrodynamics. Background Technology

[0002] With the increasing frequency of extreme rainfall events caused by climate change, the resulting floods have significantly impacted traditional hydrological stations and facilities, causing damage. Unlike general runoff processes, flood runoff processes are characterized by large water volumes, short durations, and severe meteorological conditions. Therefore, there is an urgent need for new non-contact measurement methods to monitor complete flood runoff data, including the start, peak, and end of the flood.

[0003] Currently, there are three main methods for monitoring flood runoff: hydrological station methods, hydrological models, and calculations using remote sensing data. Traditional hydrological stations construct water level-runoff relationship curves through extensive field measurements. However, with the increasing frequency of extreme rainfall events, the safety of hydrological stations during floods is compromised, and data acquisition capabilities decline or even become unavailable. Furthermore, due to natural and economic constraints, station networks are sparsely and unevenly distributed, making it difficult to obtain runoff data in many areas. Hydrological models can simulate and calculate flood runoff in a study area, including empirical models, physical conceptual models, distributed models, and semi-distributed models. Using models for long-term flow calculations requires a large amount of real-time hydrological data, including precipitation, evaporation, and runoff, as input, statistical, and validation data. For complex models, such as physically based distributed hydrological models, the uncertainty of the input data can significantly impact the accuracy of the model results. Because flood durations are short, models require substantial computing power; insufficient computing power makes it difficult for models to achieve real-time runoff monitoring, and model validation becomes problematic in areas without hydrological stations. Therefore, many regions face difficulties in calculating flood runoff using hydrological models. The development of remote sensing technology has provided a new means for monitoring flood runoff.

[0004] To acquire flood runoff data using remote sensing technologies and methods, various existing techniques utilize satellites, drones, and hydrological models to calculate and monitor flood inundation extent and runoff. For example, Chinese invention patent application number 201510092899.9 designs a method for monitoring flood inundation extent and duration using satellite remote sensing. Chinese invention patent application number 202411617924.6 designs a method for coupled monitoring and assessment of post-flood factors using drones and satellites. Chinese invention patent application number 202411317925.9 uses satellite remote sensing to monitor the water surface area before and after a flood. Chinese invention patent application number 202410248106.7 uses multi-source remote sensing data to identify post-flood factors. Chinese invention patent application number 202411383230.0 uses multi-source satellite remote sensing to obtain the boundaries and extent of flood areas. Chinese invention patent application number 202311371310.X uses satellite remote sensing data and image segmentation methods to obtain the inundation range. Chinese invention patent application number 202311564799.2 uses satellite remote sensing and Manning's formula to calculate runoff at a single cross-section during non-flood seasons. Chinese invention patent application number 202011373524.7 designs a method to measure river flow using a drone equipped with a Doppler ultrasonic flow meter.

[0005] The aforementioned existing technologies have solved the problems of obtaining flood inundation range and monitoring river runoff during non-flood periods to some extent, but they have shortcomings: 1) There is a lack of satellite-based monitoring methods for river runoff during flood processes. Currently, most related patents utilize multi-source satellites for flood inundation range monitoring and post-disaster assessment, lacking monitoring of runoff during flood processes; 2) Peak flow rates are difficult to obtain accurately. Current satellite runoff monitoring uses a single cross-section and Manning's formula, without considering the differences in water surface slope during the entire river channel movement, leading to underestimation of the calculated results; 3) The river channel roughness value does not change during the flood process. Current satellite river runoff monitoring methods... The method only sets a single roughness value, but the roughness of the river channel changes throughout the flood process, which affects the final calculation results; 4) It is not suitable for the cloudy and rainy characteristics of flood processes. Currently, most satellite-based river runoff monitoring methods use optical satellites to monitor runoff. Optical satellites have limited data available during cloudy and rainy periods, making it difficult to capture the entire flood process; 5) Continuous monitoring is costly. The equipment carried by drones to observe river flow or velocity is expensive, and the drones that can carry such equipment are also very expensive, making it difficult to conduct long-term and high-frequency river flow monitoring, especially in areas with harsh natural conditions. Summary of the Invention

[0006] The purpose of this invention is to overcome the challenge of monitoring river runoff data during flood events, and to overcome the technical deficiencies in existing technologies by proposing a method for monitoring river flood runoff that couples satellite remote sensing and hydrodynamics. This invention achieves its objective through the following technical solution:

[0007] A method for monitoring runoff during flood processes that couples satellite remote sensing and hydrodynamics includes the following steps:

[0008] Step 1: Complete UAV topographic survey of the river channel: Acquire complete UAV image data of the monitored river channel;

[0009] Step 2: Obtaining hydraulic parameters of upstream and downstream sections of the river: Determine the upstream and downstream sections of the monitored river, and measure the actual distance between the upstream and downstream sections, water depth, and flow velocity.

[0010] Step 3: Constructing a digital river cross-section: Process the UAV imagery data obtained in Step 1 to output a digital surface model and digital orthophotos, obtain basic topographic data on the water surface, and combine the cross-sectional hydraulic parameters obtained in Step 2 to fit the elevation changes above the water surface and the underwater topography, and merge them to generate a complete digital river cross-section.

[0011] Step 4: Batch extraction of water surface width from satellite remote sensing: SAR data is obtained based on satellite remote sensing data sources. The river water surface width is inverted using a hybrid pixel decomposition algorithm based on the SAR data. The threshold is then adjusted according to the measured river water surface width to extract water surface width data in batches.

[0012] Step 5: Obtain flow calculation parameters based on the digital river cross-section in Step 3 and the water surface width data obtained in Step 4: For each digital river cross-section, the specified water surface width corresponds to the corresponding water level, flow area, hydraulic radius, and wetted perimeter as flow calculation parameters;

[0013] Step Six: Remote Sensing Inversion of Complete River Channel Roughness: Land features are classified using digital orthophoto data. Different roughness coefficients are assigned to different land features according to the assignment method. The roughness of the cross-section of different water surface widths is obtained pixel-by-pixel using the remote sensing inversion results, and then calculated to form a complete, spatially continuous roughness base map of the river channel. The roughness calculation is based on the following formula:

[0014]

[0015] In the formula, n is the total roughness of the cross-section when the water surface width is RD; D is the wetted perimeter when the water surface width is RD, in meters; n j D is the roughness coefficient corresponding to the j-th pixel of the cross-section when the water surface width is RD; 像元jLet be the wetted perimeter of the j-th pixel in the cross-section when the water surface width is RD, in meters; N is the total number of pixels in the cross-section when the water surface width is RD; RD is the water surface width, in meters; w 像元 Z represents the pixel width of the drone, in meters. i and z i+1 Let be the elevations of the i-th and (i+1)-th points on the cross-section of the water passage, respectively, in meters.

[0016] Step 7: Cross-sectional flow monitoring and historical data inversion: Based on the hydraulic parameters obtained in Step 2, the flow calculation parameters obtained in Step 5, and the roughness-driven flow calculation formula obtained in Step 6, cross-sectional flow monitoring is carried out, and historical flow data is inverted. The flow calculation formula is as follows:

[0017]

[0018] Where: Q is the river flow in the studied river section, m 3 / s; K is the average water conveyance rate of the river channel, m 3 / s; S is the water surface gradient of a steady non-uniform flow; H1 is the water level at the upstream section, m; H2 is the water level at the downstream section, m; L is the distance between the upper and lower sections, m; A 上 The cross-sectional area of ​​the river section under study is m. 2 A 下 The cross-sectional area of ​​the river section under study is m. 2 ;R 上 R is the hydraulic radius of the upper section. 下 denoted as , where is the hydraulic radius of the downstream section (m); n' is the average of the total roughness of the upstream and downstream sections; and g is the acceleration due to gravity (m / s²). 2 ξ is the head loss coefficient for river contraction or diffusion; f is the kinetic energy correction coefficient.

[0019] Furthermore, the UAV image data mentioned in step one includes image data showing the river channel shape, elevation, and nearby terrain features of the target river cross-section.

[0020] Furthermore, in step two, water depth measurement is performed using continuous sonar observation; multi-point flow velocity measurement is performed using the surface flow velocity measurement method, with a flow velocity value measured at 1-meter intervals, and the instrument used is an electromagnetic current meter or a rotor current meter.

[0021] Furthermore, in step four, the formula for calculating the width of the river surface extracted by the hybrid pixel decomposition method is as follows:

[0022]

[0023] In the formula, RD is the width of the water surface in meters; x is the total number of pixels in the valley region of interest; SAR j Here is the SAR value for pixel j; SAR LT and SAR WTThese are the thresholds for land and water SAR, respectively; PA is the actual area of ​​the pixel, in meters. 2 VL is the length of the selected valley region of interest, in meters.

[0024] Furthermore, the specific formulas for the flow calculation parameters mentioned in step five are as follows:

[0025] Z = Z int (8)

[0026]

[0027] R = A / D (11)

[0028] Where Z is the water level when the river width is RD, in meters; Z int A is the elevation at the intersection of the water surface line and the cross-section, in meters; A is the flow area of ​​the upstream or downstream cross-section, in square meters. 2 Z i and Z i+1 ... i and VD i+1 R is the horizontal distance between the i-th point and the (i+1)-th point on the cross-section of the river, in meters; R is the hydraulic radius when the river width is RD, in meters.

[0029] Furthermore, in step six, different roughness coefficients are assigned to different land features, with hardened river channels set to 0.080, water bodies set to 0.030, dry land set to 0.035, irrigated land set to 0.050, forest land set to 0.083, grassland set to 0.035, and gravel set to 0.036.

[0030] The advantages and beneficial effects of this invention are:

[0031] 1) This invention revolutionizes the traditional hydrological monitoring paradigm that focuses on a single cross-section, and innovatively establishes a complete river section as the basic research unit. This systematic perspective at the river section scale can essentially capture the spatial continuity of hydrodynamic elements along the river and their dynamic evolution over time, providing fundamental support for accurately depicting the overall laws of river water distribution and transport.

[0032] 2) The method of this invention enables the rapid construction of remote sensing flow monitoring stations in natural rivers without existing stations. By simultaneously monitoring multiple sections and combining real-time dynamic capture of riverbed gradient (water surface gradient in constant non-uniform flow) and roughness changes, the accuracy of flow calculation is significantly improved, and the flow process in the complete river channel is precisely characterized. This not only optimizes the layout of existing traditional hydrological stations and effectively solves the problem of insufficient flow data in areas with no or insufficient hydrological data, but also provides long-sequence, highly reliable flow data for hydrological model parameter calibration and result verification. It better describes the flow situation during floods, making it closer to reality. Attached Figure Description

[0033] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0034] Figure 1 This is a schematic diagram illustrating the operation of an embodiment of the present invention;

[0035] Figure 2 The selected river section is used for the example study;

[0036] Figure 3 This is a water level map of the upper section on the day of the example.

[0037] Figure 4 This is a cross-sectional water level diagram for the day shown in the example.

[0038] Figure 5 The example shows the inversion results of daily runoff in 2024. Detailed Implementation

[0039] Example 1:

[0040] A method for monitoring flood runoff that couples satellite remote sensing and hydrodynamics is presented, with an implementation case study deployed at the Medog Longbaqu section in the Yarlung Tsangpo River basin. See the operational diagram below. Figure 1 Specifically, it includes the following steps:

[0041] Step 1: Unmanned Aerial Vehicle (UAV) Complete River Channel Topographic Survey: Based on the location of the selected complete river channel, such as... Figure 2 As shown, after arriving at the site, the UAV is used for field control flight, and a flight path is set according to the cross-sectional features; the UAV river channel is measured to obtain complete UAV image data of the monitored river channel; the UAV image data includes image data of the river channel shape, elevation and nearby ground features with the target river cross-section.

[0042] The monitoring river section should be located in a gentle section with stable slopes on both banks, or a standard cross section should be artificially constructed. The length between the upstream and downstream monitoring sections should not be less than 100 meters.

[0043] The initial flight altitude of the drone is set at 70-120m, and the flight path is set with the river as the center. For example, the flight range is set at 700×350m, where 700m is the length of the flight path in the direction of the river and 350m is the total width on both sides of the river.

[0044] Step 2: Obtain hydraulic parameters of upstream and downstream sections of the river: Determine the upstream and downstream monitoring sections of the river, conduct field measurements, and measure the actual distance between the upstream and downstream sections, water depth, flow velocity, slope, and roughness.

[0045] The water depth measuring instrument mainly uses sonar. The sonar detector is placed vertically on the water surface and can automatically acquire water depth data. The measurement position starts from the starting point of the cross section and collects water depth parameters continuously or every 1 meter.

[0046] Multi-point flow velocity measurement can be performed using surface flow velocity measurement methods. The instruments used can be radio wave flow meters and rotor flow meters. The measurement location starts from the beginning of the cross-section, and a flow velocity parameter is collected every 1 meter.

[0047] Step 3: Constructing a digital river cross-section: Process the UAV imagery data obtained in Step 1 to output a digital surface model and digital orthophotos, obtain basic topographic data on the water surface, and combine the cross-sectional hydraulic parameters obtained in Step 2 to fit the elevation changes above the water surface and the underwater topography, and merge them to generate a complete digital river cross-section.

[0048] Step 4: Batch Extraction of Water Surface Width from Satellite Remote Sensing: SAR data is obtained from satellite remote sensing data sources. The river water surface width is retrieved using a hybrid pixel decomposition algorithm based on the SAR data. The threshold is then adjusted according to the measured river water surface width to extract the water surface width data in batches. The calculation formula for extracting river water surface width using the hybrid pixel decomposition method is as follows:

[0049]

[0050] In the formula, RD is the width of the water surface, in meters; x is the total number of pixels in the valley region of interest; SAR j Here is the SAR value for pixel j; SAR LT and SAR WT These are the thresholds for land and water SAR, respectively; PA is the actual area of ​​the pixel, in meters. 2 VL is the length of the selected valley region of interest, in meters.

[0051] The preferred method in this embodiment is:

[0052] Training area establishment: The training area is an image region or pixel selected from the image whose land cover attributes or object characteristics are known, used for classification learning and training, and can be the region of interest; in order to perform remote sensing inversion of river flow, a total of 4 training areas need to be established: land, water body, valley, and river length.

[0053] The land training area is used to calculate the threshold of land SAR (the average SAR value of land pixels). It is necessary to select areas whose SAR values ​​are close to those of water areas but are clearly land areas based on digital orthophotos, and generate surface vector files.

[0054] The water training area is used to calculate the threshold of SAR for water bodies (the average SAR value of water body pixels). It is necessary to select water surfaces based on digital orthophotos and generate surface vector files. Water surfaces can be found in river valleys. When the water surface in a river valley is less than one pixel wide or the river is interrupted, it can also be found in nearby water bodies that exist year-round based on SAR images. Pure water surface pixels are used to represent a certain water depth.

[0055] The valley training area is a valley region used to calculate the water surface area. It is necessary to select a certain length of the river channel and the areas on both banks based on the digital orthophoto and generate a surface vector file.

[0056] The river length training area is used to calculate the average water surface width of the valley area. It is necessary to select a straight line with the same length as the valley training area based on the digital orthophoto and generate a line vector.

[0057] After the training area is established, it needs to be checked. The training area file is overlaid on the SAR data to check whether the land and water training areas are overlaid on the corresponding land and water surfaces, and whether the length of the river length training area is equal to the length of the valley training area.

[0058] Standard and Calculated Data Inputs: Standard SAR data files are SAR data with known or determined river width dates, while calculated SAR data files are SAR data with river width dates that need to be inverted, and all SAR data need to be pre-converted to a projected coordinate system.

[0059] River surface width calculation and threshold adjustment: Read in the standard SAR data, calculated SAR data and training area data file in sequence, calculate the land SAR threshold, water SAR threshold and water surface area in the standard SAR data and calculated SAR data respectively, and calculate the river surface width using formula (7); if the difference between the river surface width calculated by the standard SAR data and the river surface width measured by the UAV orthophoto on the same day is greater than the reasonable error range, the thresholds of land and water SAR need to be adjusted; recalculate the river surface width according to the adjusted thresholds of land and water SAR until the error between the standard data and the river surface width measured by the UAV orthophoto on the same day is within the reasonable range, and then calculate the river surface width using formula (7).

[0060] Step 5: Obtain flow calculation parameters based on the digital river cross section in Step 3 and the water surface width data obtained in Step 4: For each digital river cross section, the specified water surface width corresponds to the corresponding water level, flow area, hydraulic radius and wet perimeter as flow calculation parameters; the specific formulas for flow calculation parameters are as shown in equations (8) to (11).

[0061] Z = Z int (8)

[0062]

[0063] R = A / D (11)

[0064] Where Z is the water level when the river width is RD, in meters; Z int A is the elevation at the intersection of the water surface line and the cross-section, in meters; A is the flow area of ​​the upstream or downstream cross-section, in square meters. 2 Z i and Z i+1 ... i and VD i+1 R is the horizontal distance between the i-th point and the (i+1)-th point on the cross-section of the river, in meters; R is the hydraulic radius when the river width is RD, in meters.

[0065] Step 6: Remote sensing inversion of complete river channel roughness: Using the digital orthophoto data, land features are classified, and different roughness coefficients are assigned to different land features according to the assignment method. The roughness of the cross-section of different water surface widths is obtained pixel by pixel using the remote sensing inversion results, thereby forming a complete and continuous roughness base map of the river channel, and is corrected by measured roughness. The roughness is specifically calculated using formulas (1) to (3):

[0066]

[0067] In the formula, n is the total roughness of the cross-section when the water surface width is RD; D is the wetted perimeter when the water surface width is RD, in meters; n j D is the roughness coefficient corresponding to the j-th pixel of the cross-section when the water surface width is RD; 像元j Let be the wetted perimeter of the j-th pixel in the cross-section when the water surface width is RD, in meters; N is the total number of pixels in the cross-section when the water surface width is RD; RD is the water surface width, in meters; w 像元 Z represents the pixel width of the drone, in meters. i and Z i+1 Let m be the elevations of the i-th and (i+1)-th points on the cross-section of the water passage.

[0068] Different roughness coefficients are assigned to different land features, with hardened river channels set to 0.080, water bodies to 0.030, dry land to 0.035, irrigated land to 0.050, forest land to 0.083, grassland to 0.035, and gravel to 0.036.

[0069] Step 7, Cross-sectional flow monitoring and historical data inversion: Based on the hydraulic parameters obtained in Step 2, the flow calculation parameters obtained in Step 5, and the roughness-driven flow calculation formula obtained in Step 6, cross-sectional flow monitoring is carried out and historical flow data is inverted. The flow calculation formulas are as follows (4) to (6):

[0070]

[0071]

[0072] Where: Q is the river flow in the studied river section, m 3 / s; K is the average water conveyance rate of the river channel, in meters. 3 / s; S is the water surface gradient of a steady non-uniform flow; H1 is the water level at the upstream section, in meters; H2 is the water level at the downstream section, in meters; the water level values, i.e., the elevation data, are obtained by remote sensing inversion based on the relationship between the water surface width and elevation of the upstream and downstream sections; L is the distance between the upstream and downstream sections, in meters; A 上 The cross-sectional area of ​​the river section under study is m. 2 A 下 The cross-sectional area of ​​the river section under study is m. 2 ;R 上 R is the hydraulic radius of the upper section. 下 denoted as , where is the hydraulic radius of the downstream section, in meters (m); n' is the average of the total roughness of the upstream and downstream sections; g is the acceleration due to gravity, in meters per second (m / s²). 2ξ is the head loss coefficient for river contraction or diffusion. In the above formula (contraction is negative and diffusion is positive), contraction is about 0.1 to 0. For straight and gradually contracting river sections, ξ can be taken as 0. For sudden diffusion, ξ can be taken as 0.5 to 1.0. For gradual diffusion, ξ can be taken as 0.3. f is the kinetic energy correction coefficient, which is related to whether the velocity distribution on the cross section is uniform. Generally, for relatively straight river sections with small bottom slopes and relatively regular cross sections, its value is between 1.05 and 1.15, and f can be taken as 1.0.

[0073] The following is the relevant data for this embodiment:

[0074] Table 1 is the cross-sectional record table for the day 2025032703;

[0075] Table 2 shows the calculation results of cross-sectional parameters for the day 2025032703;

[0076] Table 3 shows the results of the flow calculation and inversion in 2024.

[0077] Table 1

[0078]

[0079] Key calculation results of cross-sectional parameters for the day:

[0080] Table 2

[0081] 2025032703 Digital River Channel Calculation results unit Upper section water surface width 16.77 m Lower section water surface width 12.01 m Gradient (water surface gradient) 0.00399 / Roughness 0.029 / Upper section water level 2918.2 m Lower section water level 2918.0005 m Average water depth at upper section 0.433 m Average water depth at lower cross section 0.699 m Upper cross-sectional velocity 0.7832 m / s Lower cross-sectional velocity 1.0391 m / s River flow 3.58 <![CDATA[m 3 / s <!-- 7 -->]]> Flow rate adjustment factor 1 / Water surface width error 2 ％ Cross-sectional smoothing factor 1 / Upper cross-sectional water flow area 4.5619 <![CDATA[m 2 ]]> Lower cross-sectional water flow area 3.2879 <![CDATA[m 2 ]]> upper section wet perimeter 17.4489 m Lower section wet perimeter 12.4787 m upper section hydraulic radius 0.2614 m Hydraulic radius of lower section 0.2634 m Kinetic energy correction coefficient 1.03 / Head loss coefficient 0.05 / Spacing between upper and lower sections 50 m

[0082] Table 3

[0083]

[0084]

[0085] In this calculation case, when the hydraulic gradient (gradient) of the Manning formula is fixed at 0.0039955422, the analysis found that during the high-water season when the river surface widens and the flow rate increases significantly (such as January 8th and 13th, 2024), the calculation results of the Manning formula have the limitation that the flow rate is systematically underestimated (i.e., "the flow rate does not increase"). This is mainly because a fixed gradient cannot fully reflect the dynamic changes in the actual hydraulic gradient of the river channel. In contrast, the algorithm of this invention, through its core mechanism—real-time dynamic updating of the actual hydraulic gradient between the upper and lower sections (Formula (6))—combined with kinetic energy correction for uneven flow velocity distribution in the section and detailed consideration of head loss caused by bends, contraction / expansion, etc., significantly overcomes the shortcomings of the fixed gradient method. This dynamic adaptability and comprehensive processing of complex hydraulic elements enable the method of this invention to capture the spatiotemporal variation characteristics of river flow more realistically and sensitively, thereby significantly improving the accuracy and reliability of the simulation of the entire flow process line.

[0086] Finally, it should be noted that the above description is only used to illustrate the technical solutions of the present invention and is not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention.

Claims

1. A method for monitoring runoff during flood processes that couples satellite remote sensing and hydrodynamics, characterized in that: Includes the following steps: Step 1: Complete UAV topographic survey of the river channel: Acquire complete UAV image data of the monitored river channel; Step 2: Obtaining hydraulic parameters of upstream and downstream sections of the river: Determine the upstream and downstream sections of the monitored river, and measure the actual distance between the upstream and downstream sections, water depth, and flow velocity. Step 3: Constructing a digital river cross-section: Process the UAV imagery data obtained in Step 1 to output a digital surface model and digital orthophotos, obtain basic topographic data on the water surface, and combine the cross-sectional hydraulic parameters obtained in Step 2 to fit the elevation changes above the water surface and the underwater topography, and merge them to generate a complete digital river cross-section. Step 4: Batch extraction of water surface width from satellite remote sensing: SAR data is obtained based on satellite remote sensing data sources. The river water surface width is inverted using a hybrid pixel decomposition algorithm based on the SAR data. The threshold is then adjusted according to the measured river water surface width to extract water surface width data in batches. Step 5: Obtain flow calculation parameters based on the digital river cross-section in Step 3 and the water surface width data obtained in Step 4: For each digital river cross-section, the specified water surface width corresponds to the corresponding water level, flow area, hydraulic radius, and wetted perimeter as flow calculation parameters; Step Six: Remote Sensing Inversion of Complete River Channel Roughness: Land features are classified using digital orthophoto data. Different roughness coefficients are assigned to different land features according to the assignment method. The roughness of the cross-section of different water surface widths is obtained pixel-by-pixel using the remote sensing inversion results, and then calculated to form a complete, spatially continuous roughness base map of the river channel. The roughness calculation is based on the following formula: In the formula, n is the total roughness of the cross-section when the water surface width is RD; D is the wetted perimeter when the water surface width is RD, in meters; n j D is the roughness coefficient corresponding to the j-th pixel of the cross-section when the water surface width is RD; 像元j Let be the wetted perimeter of the j-th pixel in the cross-section when the water surface width is RD, in meters; N is the total number of pixels in the cross-section when the water surface width is RD; RD is the water surface width, in meters; w 像元 Z represents the pixel width of the drone, in meters. i and Z i+1 Let be the elevations of the i-th and (i+1)-th points on the cross-section of the water passage, respectively, in meters. Step 7: Cross-sectional flow monitoring and historical data inversion: Based on the hydraulic parameters obtained in Step 2, the flow calculation parameters obtained in Step 5, and the roughness-driven flow calculation formula obtained in Step 6, cross-sectional flow monitoring is carried out, and historical flow data is inverted. The flow calculation formula is as follows: Where: Q is the river flow in the studied river section, m 3 / s; K is the average water conveyance rate of the river channel, m 3 / s; S is the water surface gradient of a steady non-uniform flow; H1 is the water level at the upstream section, m; H2 is the water level at the downstream section, m; L is the distance between the upper and lower sections, m; A 上 The cross-sectional area of ​​the river section under study is m. 2 A 下 The cross-sectional area of ​​the river section under study is m. 2 ;R 上 R is the hydraulic radius of the upper section. 下 denoted as , where is the hydraulic radius of the downstream section (m); n' is the average of the total roughness of the upstream and downstream sections; and g is the acceleration due to gravity (m / s²). 2 ξ is the head loss coefficient for river contraction or diffusion; f is the kinetic energy correction coefficient.

2. The flood process runoff monitoring method coupled with satellite remote sensing and hydrodynamics according to claim 1, characterized in that: The UAV image data mentioned in step one includes image data showing the shape, elevation, and nearby features of the target river cross-section.

3. The flood process runoff monitoring method coupled with satellite remote sensing and hydrodynamics according to claim 1, characterized in that: In step two, water depth measurement is performed using continuous sonar observation; multi-point flow velocity measurement is performed using the surface flow velocity measurement method, with a flow velocity value measured at 1-meter intervals, and the instrument used is an electromagnetic current meter or a rotor current meter.

4. The flood process runoff monitoring method coupled with satellite remote sensing and hydrodynamics according to claim 1, characterized in that: In step four, the formula for calculating the width of the river surface using the hybrid pixel decomposition method is as follows: In the formula, RD is the width of the water surface in meters; x is the total number of pixels in the valley region of interest; SAR j Here is the SAR value for pixel j; SAR LT and SAR WT These are the thresholds for land and water SAR, respectively; PA is the actual area of ​​the pixel, in meters. 2 VL is the length of the selected valley region of interest, in meters.

5. The flood process runoff monitoring method coupled with satellite remote sensing and hydrodynamics according to claim 1, characterized in that: The specific formulas for the flow calculation parameters mentioned in step five are as follows: Z=Z int (8) R = A / D (11) Where Z is the water level when the river width is RD, in meters; Z int A is the elevation at the intersection of the water surface line and the cross-section, in meters; A is the flow area of ​​the upstream or downstream cross-section, in square meters. 2 Z i and Z i+1 ... i and VD i+1 R is the horizontal distance between the i-th point and the (i+1)-th point on the cross-section of the river, in meters; R is the hydraulic radius when the river width is RD, in meters.

6. The flood process runoff monitoring method coupled with satellite remote sensing and hydrodynamics according to claim 1, characterized in that: In step six, different roughness coefficients are assigned to different land features, with hardened river channels set to 0.080, water bodies set to 0.030, dry land set to 0.035, irrigated land set to 0.050, forest land set to 0.083, grassland set to 0.035, and gravel set to 0.036.

Images