A flood prevention early warning and monitoring method for a water conservancy hub based on radar and satellite data

Abstract

The present application belongs to the technical field of flood prevention early warning monitoring, and relates to a water conservancy hub flood prevention early warning monitoring method based on radar and satellite data. The present application identifies cloud cluster active areas by fusing satellite cloud top brightness temperature and radar echo data, generates multi-level initial early warning signals according to the change trend of the overlapping area, dynamically adjusts the collection frequency according to the early warning level, calculates the rain intensity cumulative amount and water level rising gradient, deduces the dam front water level prediction sequence and extracts the peak value, generates the updated early warning signal in combination with the predicted peak water level, the current water level and the flood control water level relationship, sets the level comparison to adjust the collection frequency, and takes the updated signal as the next round of input to realize closed-loop iteration. The present application solves the problems that the traditional method cannot lock the high-risk area in advance, the monitoring frequency is fixed, the future water level and the early warning level cannot be predicted, and the early warning level frequently switches, and realizes the proactive early warning before the flood, the dynamic optimization of monitoring resources, the advanced water level prediction and the stable update of early warning.

Description

Technical Field

[0001] This invention belongs to the field of flood control early warning and monitoring technology, and relates to a flood control early warning and monitoring method for water conservancy projects based on radar and satellite data. Background Technology

[0002] Water conservancy projects play a crucial role in flood control, disaster reduction, water resource allocation, and power generation, and their safe operation is directly related to the safety of people's lives and property downstream. During the flood season, real-time monitoring and early warning of the meteorological and hydrological conditions of the basin where the project is located are the core links to ensure dam safety and achieve scientific scheduling.

[0003] However, in practical applications, the existing technology has the following specific problems:

[0004] First, traditional methods often only initiate a response after rainfall occurs or reservoir water levels rise. Although some systems have incorporated meteorological radar and satellite remote sensing data, they lack joint analysis of the trends in both data, making it impossible to identify high-risk areas before precipitation systems affect the basin. This results in missed opportunities for early dispatch, and the early warning response remains passive and delayed.

[0005] Secondly, existing monitoring methods mostly use a constant sampling period. During periods of low risk, high-frequency sampling results in data redundancy and energy waste. During periods of heightened risk, low-frequency sampling fails to capture key process details such as rapid cloud evolution and sudden changes in rainfall intensity. This leads to the data foundation for subsequent risk assessment lagging behind the development of the flood situation, making it difficult to balance monitoring efficiency and accuracy.

[0006] Furthermore, current early warnings are mostly based on comparing real-time water levels with warning lines, which is a passive response. They do not combine upstream rainfall processes with water level change trends to predict future water level changes, making it impossible to predict whether the safety threshold will be exceeded. This weakens the initiative of pre-release and reservoir emptying, and the early warning level cannot be dynamically adjusted according to future water level predictions, resulting in insufficient foresight in early warnings.

[0007] Finally, the early warning level update is based on only a single result. When the data fluctuates at the critical value, the level switches frequently, causing confusion in dispatch instructions. Furthermore, there is a lack of a continuous confirmation mechanism when the risk decreases, and hastily reducing the monitoring intensity may lead to the loss of secondary disasters. Summary of the Invention

[0008] In view of this, in order to solve the problems mentioned in the background technology, a flood control early warning and monitoring method for water conservancy projects based on radar and satellite data is proposed.

[0009] The objective of this invention can be achieved through the following technical solution: a flood control early warning monitoring method for water conservancy hubs based on radar and satellite data, comprising: acquiring satellite cloud top brightness temperature and radar echo data at a basic acquisition frequency to identify active cloud areas, and overlapping with the water conservancy hub control basin; if the rate of change of the area ratio of the overlapping area is greater than zero and the change in cloud coldness is less than zero, then generating multi-level initial early warning signals.

[0010] The acquisition frequency is adjusted according to the initial warning signal level, and the rainfall intensity time series and the water level time series in front of the dam are acquired according to the adjusted frequency.

[0011] Using the overlapping area determined by the early warning signal as the range, the rainfall intensity is spatially and temporally accumulated to obtain the rainfall intensity accumulation sequence. The water level rise gradient sequence is obtained by differentiating the water level in front of the dam. The predicted water level in front of the dam is derived, and the predicted peak water level is extracted from it.

[0012] By combining the predicted peak water level with the relationship between the current water level and the flood control limit water level, an updated early warning signal is generated. The updated level is compared with the previous level to adjust the acquisition frequency, and the updated early warning signal is used as the input for the next round of acquisition frequency adjustment.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention identifies the active area of ​​cloud clusters and overlaps with the watershed by fusing satellite cloud top brightness temperature and radar echo intensity data, and generates an initial warning signal based on the rate of change of the area ratio of the overlapping area and the change in cloud cluster coldness. This solves the problem that traditional technologies lack joint analysis of multi-source data trends and cannot lock high-risk areas before the impact of precipitation. By quantifying the spatial relationship between cloud cluster development and watershed, the risk can be captured before rainfall occurs, improving the timeliness and foresight of the warning, and buying time for the early scheduling of the hub.

[0014] (2) This invention dynamically adjusts the acquisition frequency according to the initial warning signal level. When the risk is low, the basic frequency is maintained, and when the risk is high, the radar, rainfall intensity, and water level monitoring frequencies are increased. This solves the problem that existing monitoring uses a constant sampling period and cannot capture key process details in high-risk periods. By matching the acquisition density with the risk level in a graded manner, the optimal allocation of monitoring resources is achieved, avoiding data lag in low-frequency acquisition and reducing energy consumption in high-frequency acquisition, while effectively balancing monitoring efficiency and accuracy for stable operation.

[0015] (3) This invention calculates the cumulative rainfall intensity and water level rise gradient within the overlapping area of ​​the early warning system, extrapolates the predicted water level sequence in front of the dam, and extracts the predicted peak water level. This solves the problem that current early warning systems rely on comparing real-time water levels with warning lines and do not incorporate upstream rainfall to extrapolate future water levels, thus failing to predict when the safety threshold will be exceeded. By extrapolating future water levels through the model, the early warning level is dynamically adjusted based on future water level predictions, effectively enhancing the initiative of pre-release for reservoir emptying and significantly improving the foresight of the early warning system.

[0016] (4) This invention dynamically adjusts the collection frequency by comparing the updated level with the previous level. This solves the problem that the early warning level update is based only on a single result, and the frequent switching of levels is caused by fluctuations in the data threshold, and there is a lack of a continuous confirmation mechanism when the risk decreases. By setting a confirmation threshold for each monitoring round, the collection frequency is adjusted only after the risk has stabilized and decreased, avoiding frequent changes in the early warning level caused by fluctuations in the data threshold, and ensuring the continuity and stability of flood control scheduling. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the steps of a flood control early warning and monitoring method for water conservancy hubs based on radar and satellite data, as described in this invention.

[0019] Figure 2 This is a flowchart illustrating the specific method for identifying active areas of cloud clusters in this invention.

[0020] Figure 3 This is a flowchart of the method for generating multi-level initial warning signals in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0023] The following description, in conjunction with the accompanying drawings, details a specific scheme for a flood control early warning and monitoring method for water conservancy hubs based on radar and satellite data provided by the present invention.

[0024] Please see Figure 1 As shown, the present invention provides a flood control early warning and monitoring method for water conservancy projects based on radar and satellite data, the method comprising S1 to S4.

[0025] S1. Use the basic acquisition frequency to acquire satellite cloud top brightness temperature and radar echo data to identify the active area of ​​the cloud cluster, and overlap it with the water conservancy hub control basin. If the rate of change of the area ratio of the overlapping area is greater than zero and the change in cloud cluster coldness is less than zero, then generate a multi-level initial warning signal.

[0026] Due to the vast upstream basin of the hydropower project and the sparse distribution of ground monitoring stations, it is difficult to capture the formation and movement trajectory of large-scale precipitation clouds in a timely manner. Satellite cloud top brightness temperature data can characterize the vertical development trend of clouds, while radar echo data can characterize the precipitation intensity distribution within clouds. Therefore, in order to identify strong precipitation clouds that may move into the basin controlled by the project at an early stage, it is necessary to integrate satellite brightness temperature and radar echo data and use spatiotemporal matching analysis to identify active areas of clouds that are in the development stage and overlap with the basin.

[0027] In one specific embodiment, satellite cloud top brightness temperature and radar echo data are acquired at a base acquisition frequency to identify active cloud areas. The base frequency can be set to a satellite scan every 30 minutes and a radar scan every 15 minutes; the implementer can set this frequency according to specific circumstances.

[0028] Please see Figure 2 As shown, the specific method for identifying active cloud areas is as follows: Satellite cloud top brightness temperature data is mapped into a gridded brightness temperature field. Each grid is traversed; if the brightness temperature value at the current acquisition time is less than the brightness temperature value at the previous acquisition time, and this state continues for N consecutive acquisition times, it indicates that the cloud top height is rising and convective activity is increasing, and this grid is marked as a brightness temperature decreasing grid. The satellite cloud top brightness temperature data is downloaded in real-time through a meteorological satellite data receiving interface.

[0029] The radar echo data is mapped into a gridded echo field. Each grid is traversed, and if the echo intensity value at the current acquisition time is greater than the echo intensity value at the previous acquisition time, and this state continues for N consecutive acquisition times, it indicates that precipitation particles within the cloud are increasing and the echo intensity is strengthening; this is then marked as an echo-rising grid. The radar echo data is acquired in real time through a radar detection system.

[0030] The N is usually set to an integer between 2 and 5. In this embodiment, N=3 is preferred. The implementer can set it according to the specific situation.

[0031] The eight-neighbor connected regions of the brightness temperature decreasing grid and the echo increasing grid are extracted, and the spatial intersection region of the two is taken as the cloud active area. This area represents a high-risk precipitation area that simultaneously has cloud top cooling and echo enhancement characteristics.

[0032] Furthermore, since the development and evolution of cloud clusters can only have a substantial impact on the flood control safety of water conservancy hubs when there is spatial overlap between the active area of ​​cloud clusters and the control basin of water conservancy hubs, it is necessary to conduct spatial overlay analysis of the two.

[0033] Specifically, please refer to Figure 3 As shown, by statistically analyzing the area ratio of overlapping regions and calculating its first derivative with respect to time as the rate of area change, the expansion trend of the cloud cluster's influence range within the watershed is characterized. At the same time, the first derivative of the average cloud top brightness temperature with respect to time within the active cloud cluster area is calculated as the change in coldness to quantify the evolution characteristics of the cloud cluster's convection intensity.

[0034] The method for obtaining the area ratio of the overlapping region is as follows: count the number of spatially overlapping grids between the active cloud area and the water conservancy hub control basin, and the total number of grids within the water conservancy hub control basin. The ratio of the number of spatially overlapping grids to the total number of grids is taken as the area ratio of the overlapping region.

[0035] When the rate of area change is greater than zero and the change in coldness is less than zero, it indicates that the cloud cover area of ​​the watershed is expanding and convection is enhanced. A threshold sequence of the rate of area change and a threshold sequence of the change in coldness are constructed. The threshold sequence is based on the cumulative frequency distribution curve of the corresponding parameters in historical heavy rainfall events, and specific percentiles are selected as interval boundaries. For example, 50%, 80% and 95% percentiles are selected to divide the low, medium and high level intervals.

[0036] The calculated rate of change of area and the change of coldness fall into the level intervals divided by the threshold sequence, and a warning signal is generated according to the combination of level intervals: if both fall into the high level interval, a red warning signal is generated; if either falls into the high level interval or both fall into the medium level interval, an orange warning signal is generated; if either falls into the medium level interval, a yellow warning signal is generated; if both fall into the low level interval, a blue warning signal is generated.

[0037] When the rate of change of area is less than or equal to zero and the change in temperature is greater than or equal to zero, a blue-level initial warning signal is generated by default, or the warning signal output in the previous round is maintained as the initial warning signal for this round.

[0038] S2. Adjust the acquisition frequency according to the initial warning signal level, and acquire the rainfall intensity time series and the water level time series in front of the dam according to the adjusted frequency.

[0039] Considering the differences in resource consumption and data accuracy requirements for flood control monitoring under different warning levels, fixed-frequency data acquisition may lead to resource waste during low-risk periods or insufficient data during high-risk periods. Therefore, dynamic optimization of monitoring resources is needed. Thus, by adjusting the acquisition frequency according to the initial warning signal level, rainfall intensity time-series and dam-front water level time-series data are obtained based on the adjusted frequency.

[0040] In one specific embodiment, the adjustment of the acquisition frequency according to the initial warning signal level is as follows: When a blue warning is issued, it indicates that the expansion trend of the active cloud area and the overlapping range of the watershed is weak and the convection development is slow, and the flood control risk is within a controllable range. At this time, the basic acquisition frequency of satellites and radars is maintained.

[0041] A yellow alert indicates that the area of ​​cloud activity overlaps significantly with the watershed or that convection is accelerating, posing a potential risk of rainfall. At this time, the radar acquisition frequency is increased to the first multiple of the radar scanning frequency in the basic acquisition frequency.

[0042] An orange alert indicates that the area of ​​cloud activity overlaps rapidly with the watershed and the intensity of convection is significantly enhanced, with a high probability of rainfall. At this time, ground-based rainfall intensity stations are activated to monitor the rainfall intensity in the watershed in real time, and the rainfall intensity collection frequency is increased to the second multiple of the radar scanning frequency in the basic collection frequency.

[0043] A red alert indicates that the area of ​​cloud activity overlaps rapidly with the watershed and the intensity of convection is drastically increasing, indicating that heavy rainfall is about to occur or has already occurred. At this time, the upstream section of the hub and the water level monitoring in front of the dam are activated to obtain inflow and water level data in real time. At the same time, the water level acquisition frequency in front of the dam is increased to the third multiple of the radar scanning frequency in the basic acquisition frequency.

[0044] The first, second, and third multiples are set to 2, 4, and 6 times respectively, and the implementer can set them according to the specific circumstances.

[0045] S3. Using the overlapping area determined by the early warning signal as the range, the rainfall intensity is spatially and temporally accumulated to obtain the rainfall intensity accumulation sequence. The water level difference in front of the dam is used to obtain the water level rise gradient sequence. The water level prediction sequence in front of the dam is derived, and the predicted peak water level is extracted from it.

[0046] Considering the temporal and spatial lag of the rainfall runoff confluence process, and the fact that the water level in front of the dam is affected by both the inflow and the current water level gradient, it is impossible to predict the future flood peak directly using the current water level. Therefore, it is necessary to combine the cumulative rainfall intensity and the water level rise gradient to realize the water level in front of the dam through a hydrological and hydraulic model.

[0047] In one specific embodiment, the method for obtaining the dam front water level prediction sequence is as follows: traverse all grids within the overlapping area, accumulate the rainfall intensity grid values ​​of each acquisition cycle to obtain the spatial rainfall intensity sum, and then accumulate and sum the spatial rainfall intensity sum according to the time series to obtain the rainfall intensity accumulation sequence. This sequence represents the total accumulated rainfall within the overlapping area of ​​the cloud cluster active area and the watershed from the start of rainfall to the current time, reflecting the watershed's runoff potential and the changing trend of future inflow.

[0048] Iterate through the K consecutive sampling times preceding the current sampling time, calculate the difference in water level in front of the dam between each sampling time and the previous sampling time, and divide by the corresponding time interval to obtain the water level rise gradient sequence. This sequence characterizes the instantaneous rate of change of water level in front of the dam and reflects the current storage and release status of the reservoir area. For example, K can be set to 4.

[0049] Input the cumulative rainfall intensity sequence into the runoff generation and confluence calculation model to output the inflow prediction sequence for multiple consecutive future collection times starting from the current collection time; input the inflow prediction sequence and the water level rise gradient sequence into the reservoir flood control calculation model to output the dam front water level prediction sequence for multiple consecutive future collection times starting from the current collection time.

[0050] The runoff calculation model adopts the Xin'anjiang model to convert areal rainfall into inflow. The spatial grid resolution of the Xin'anjiang model is consistent with the grid resolution of the radar echo data, for example, 1km×1km, and the time step is synchronized with the adjusted acquisition frequency.

[0051] The reservoir flood control calculation model uses a water balance equation based on the reservoir capacity curve and discharge curve to simulate the reservoir regulation process. The model parameters are set based on the historical flood data of the basin over the past 5 years.

[0052] The multiple continuous data collection times can be set to 6 to 12 future data collection cycles, corresponding to a warning and foresight period of 3 to 6 hours in the future.

[0053] Subsequently, all water level values ​​included in the dam front water level prediction sequence are iterated through, and the maximum value among all water level values ​​is selected as the predicted peak water level, which represents the highest water level that may be reached in front of the dam during the future early warning period.

[0054] S4. Based on the predicted peak water level and the relationship between the current water level and the flood control limit water level, an updated early warning signal is generated. The updated level is compared with the previous level to adjust the acquisition frequency, and the updated early warning signal is used as the input for the next round of acquisition frequency adjustment.

[0055] Given that the flood control limit water level of a water conservancy project is the core threshold for flood control safety, the relative relationship between the predicted peak water level, the flood control limit water level, and the current water level directly reflects the flood control risk level of the project. The initial warning signal is generated solely based on cloud characteristics and does not consider the actual water storage status of the reservoir or future inflow conditions. Therefore, it must be updated in conjunction with water level prediction results to ensure that the warning signal accurately reflects the comprehensive risks faced by the project.

[0056] Meanwhile, the data collection frequency needs to be dynamically adjusted according to changes in the warning level, so as to increase monitoring intensity when the risk increases and conserve resources when the risk decreases. Furthermore, to avoid frequent rises and falls in warning levels due to fluctuations in a single prediction, which could cause instability in the monitoring system, a confirmation mechanism needs to be set up for warning level declines. Finally, by using the updated warning signal as input for the next round of monitoring, a closed-loop iteration of the entire warning monitoring process is achieved.

[0057] In one specific embodiment, the step of generating an updated early warning signal by combining the predicted peak water level and the relationship between the current water level and the flood control limit water level is as follows: if the predicted peak water level is higher than or equal to the flood control limit water level, it indicates that the predicted flood peak will break through the safety defense line and the hub faces an emergency flood control risk, and then an updated early warning signal at the red level is generated.

[0058] If the predicted peak water level is lower than the flood limit water level but higher than or equal to the current water level, it indicates that the water level is still rising. Although it has not exceeded the flood limit water level, it has shown an upward trend and there is a potential risk of exceeding the flood limit. In this case, an updated warning signal at the orange level will be generated.

[0059] If the predicted peak water level is lower than the current water level and the current water level is higher than or equal to the flood limit water level, it means that although the water level has exceeded the flood limit water level, it is receding and the risk is in a decreasing phase. However, we still need to be vigilant about the continued pressure brought by the higher water level, and an updated warning signal at the yellow level will be generated.

[0060] If the predicted peak water level is lower than the current water level and the current water level is lower than the flood control limit, it indicates that the water level is stable or declining and within a safe range, with no immediate risk of flooding. In this case, an updated warning signal at the blue level will be generated.

[0061] Furthermore, since the rise and fall of the warning level directly reflects changes in the risk situation, it is necessary to make differentiated adjustments to the monitoring intensity accordingly. Therefore, the updated level will be compared with the previous level, and the following rules will be implemented:

[0062] If the update level is higher than the previous level, it indicates an escalation of the risk situation. In this case, the data collection frequency will be increased immediately to ensure high-frequency data capture and real-time response in emergency situations.

[0063] If the update level is lower than the previous level, it indicates that the risk situation has eased. The number of monitoring rounds at this update level is counted to prevent the frequency from being mistakenly reduced due to data fluctuations. When the number of monitoring rounds reaches the preset confirmation threshold, the frequency is adjusted to the collection frequency corresponding to the update level, indicating that the risk has been confirmed to have been steadily reduced, and monitoring resources can be optimized.

[0064] If the preset confirmation threshold is not reached, it indicates that the risk reduction trend may be fluctuating or not yet stable, and the collection frequency of the previous round remains unchanged. The preset confirmation threshold can be set to 3 to 5 monitoring rounds, which the implementer can adjust according to specific circumstances. A monitoring round refers to a complete collection and processing cycle, the length of which is synchronized with the current collection frequency.

[0065] Furthermore, to achieve a closed-loop iteration of early warning and monitoring, the results of this round of updates need to be used as the initial conditions for the next round of monitoring, specifically including:

[0066] The updated warning signal generated in this round will be used as the initial warning signal for the next round of monitoring, indicating that the next round will start with the initial configuration of the sampling frequency from this level. The updated level in this round will be used as the previous level for the next round of monitoring, for level comparison at the end of the next round.

[0067] In the next round of monitoring, the frequency adjustment step is performed based on the initial warning signal, and the above process is repeated to achieve a closed-loop iteration of the entire process from data acquisition to warning generation and then to frequency adjustment.

[0068] In summary, this invention utilizes satellite and radar collaborative monitoring to acquire cloud top brightness temperature and echo intensity data at a basic acquisition frequency. It identifies active cloud areas and performs spatial overlap analysis with the water conservancy hub's controlled basin. When the rate of change of the overlapping area ratio is greater than zero and the change in cloud coldness is less than zero, multi-level initial warning signals are generated. The acquisition frequency is dynamically adjusted based on the initial warning signal level to acquire rainfall intensity time series and dam front water level time series data at the adjusted frequency. Using the overlapping area determined by the current warning signal as the range, the rainfall intensity is spatially and temporally accumulated to obtain a rainfall intensity accumulation sequence. The dam front water level difference is used to obtain a water level rise gradient sequence, which is then used to deduce the dam front water level prediction sequence and extract the predicted peak water level.

[0069] Simultaneously, by combining the predicted peak water level with the comparison between the current water level and the flood control limit water level, an updated early warning signal is generated. The updated level is compared with the previous level, and strategies such as increasing the sampling frequency, adjusting the sampling frequency based on the confirmation threshold, or keeping it unchanged are implemented according to different situations of rising, falling, or maintaining. The updated early warning signal is used as the input for the next round of sampling frequency adjustment, realizing a closed-loop iteration of the entire process from cloud cluster identification to water level prediction to early warning update and frequency feedback.

[0070] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0071] Those skilled in the art will recognize that the algorithmic steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0072] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0073] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0074] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for flood control early warning and monitoring of water conservancy projects based on radar and satellite data, characterized in that, include: Satellite cloud top brightness temperature and radar echo data are acquired using the basic acquisition frequency to identify active cloud areas, which overlap with the water conservancy hub control basin. If the rate of change of the area ratio of the overlapping area is greater than zero and the change in cloud coldness is less than zero, multi-level initial warning signals are generated. The acquisition frequency is adjusted according to the initial warning signal level, and the rainfall intensity time series and the water level time series in front of the dam are acquired according to the adjusted frequency; Using the overlapping area determined by the early warning signal as the range, the rainfall intensity is spatially and temporally accumulated to obtain the rainfall intensity accumulation sequence. The water level rise gradient sequence is obtained by differentiating the water level in front of the dam. The water level prediction sequence in front of the dam is derived, and the predicted peak water level is extracted from it. By combining the predicted peak water level with the relationship between the current water level and the flood control limit water level, an updated early warning signal is generated. The updated level is compared with the previous level to adjust the acquisition frequency, and the updated early warning signal is used as the input for the next round of acquisition frequency adjustment.

2. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The specific method for identifying active areas of cloud clusters is as follows: The satellite cloud top brightness temperature data is mapped into a gridded brightness temperature field. Each grid is traversed. If the brightness temperature value at the current acquisition time is less than the brightness temperature value at the previous acquisition time, and the current state continues for N consecutive acquisition times, it is marked as a brightness temperature decreasing grid. The radar echo data is mapped into a gridded echo field. Each grid is traversed. If the echo intensity value at the current acquisition time is greater than the echo intensity value at the previous acquisition time, and the current state continues for N consecutive acquisition times, it is marked as an echo rising grid. The eight-neighbor connected regions of the brightness temperature decreasing grid and the echo increasing grid are extracted, and the spatial intersection region of the two is taken as the active area of ​​the cloud.

3. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The method for obtaining the area ratio of the overlapping region is as follows: The number of spatially overlapping grids between the active cloud area and the water conservancy hub control basin, as well as the total number of grids within the water conservancy hub control basin, are statistically analyzed. The ratio of the number of spatially overlapping grids to the total number of grids is used as the area ratio of the overlapping region.

4. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The method for generating the multi-level initial warning signal is as follows: The first derivative of the area ratio of the overlapping region with respect to time is used as the rate of area change. The first derivative of the average cloud top brightness temperature with respect to time within the active cloud region is calculated as the change in cooling. When the rate of change of area is greater than zero and the change in temperature is less than zero, a multi-level initial warning signal of blue, yellow, orange and red is mapped to the area change rate threshold sequence and the temperature change threshold sequence, respectively.

5. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 4, characterized in that, The specific details of adjusting the acquisition frequency based on the initial warning signal level are as follows: During a blue alert, maintain the basic data acquisition frequencies of satellites and radar; When a yellow alert is issued, the radar acquisition frequency is increased to the first multiple of the radar scanning frequency in the basic acquisition frequency; During an orange alert, ground-based rainfall intensity monitoring stations are activated, and the rainfall intensity collection frequency is increased to the second multiple of the radar scanning frequency in the basic collection frequency. When a red alert is issued, monitoring of the upstream section of the hub and the water level in front of the dam will be activated, and the water level acquisition frequency in front of the dam will be increased to the third multiple of the radar scanning frequency in the basic acquisition frequency.

6. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The method for obtaining the predicted water level sequence in front of the dam is as follows: Traverse all grids within the overlapping area, accumulate the rainfall intensity grid values ​​of each acquisition cycle to obtain the spatial rainfall intensity sum, and then accumulate and sum the spatial rainfall intensity sum according to the time series to obtain the rainfall intensity cumulative amount sequence. Iterate through the K consecutive sampling times before the current sampling time, calculate the difference in water level in front of the dam between each sampling time and the previous sampling time, and divide by the corresponding time interval to obtain the water level rise gradient sequence. Input the cumulative rainfall intensity sequence into the runoff calculation model and output the inflow prediction sequence corresponding to multiple consecutive future collection times starting from the current collection time. The inflow prediction sequence and the water level rise gradient sequence are input together into the reservoir flood control calculation model, and the output is the water level prediction sequence in front of the dam corresponding to multiple consecutive future collection times starting from the current collection time.

7. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The method for extracting the predicted peak water level is as follows: Iterate through all water level values ​​in the predicted water level sequence in front of the dam, and select the maximum value among all water level values ​​as the predicted peak water level.

8. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The updated early warning signal is generated by combining the predicted peak water level with the relationship between the current water level and the flood control limit water level. Specifically: If the predicted peak water level is higher than or equal to the flood control limit water level, an updated warning signal at the red level will be generated. If the predicted peak water level is lower than the flood control limit water level but higher than or equal to the current water level, an updated warning signal at the orange level will be generated. If the predicted peak water level is lower than the current water level and the current water level is higher than or equal to the flood control limit water level, an updated warning signal at the yellow level will be generated. If the predicted peak water level is lower than the current water level and the current water level is lower than the flood limit water level, an updated warning signal at the blue level will be generated.

9. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The step of adjusting the collection frequency by comparing the updated level with the previous level includes: If the update level is higher than the previous level, immediately increase the collection frequency; If the update level is lower than the previous level, count the number of monitoring rounds that the update level continues to be; When the number of monitoring rounds reaches the preset confirmation threshold, the frequency will be adjusted to the collection frequency corresponding to the update level. If the preset confirmation threshold is not reached, the collection frequency of the previous round will remain unchanged.

10. The flood control early warning and monitoring method for water conservancy projects based on radar and satellite data as described in claim 1, characterized in that, The updated early warning signal serves as the input for the next round of frequency adjustment, including: The updated warning signal generated in this round will be used as the initial warning signal for the next round of monitoring; The update level of this round will be used as the previous level for the next round of monitoring; In the next round of monitoring, the sampling frequency will be adjusted based on the initial warning signal to achieve closed-loop iteration.

Images