A quantitative prediction method and system for extended-range precipitation in reservoir areas
By establishing a global model-driven mesoscale model and a multi-index model for reservoir basin precipitation forecasting, the problem of inaccurate reservoir basin precipitation forecasting has been solved, enabling quantitative forecasting for the next 7-30 days and supporting the safe and economical operation of reservoirs and the security of power supply.
Patent Information
- Application Number
- CN202211115851.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-14
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2042-09-14
AI Technical Summary
Existing technologies cannot accurately predict rainfall down to the small-scale reservoir basin, posing challenges to reservoir operation and flood control, especially in cases of frequent extreme rainfall, making it impossible to support long-term operation plans and maintenance schedules.
By collecting information on reservoir basins and circulation systems, using global models to drive mesoscale models, and combining a multi-index model system for reservoir basin precipitation prediction, a statistical relationship equation between model output circulation factors and areal rainfall is established to achieve quantitative predictions for the next 7-30 days.
It enables rainfall forecasting in mesoscale reservoir basins, supports the safe and economical operation of reservoirs during the flood season, and ensures the safety of social power supply.
Smart Images

Figure CN115423194B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power grid disaster prevention and mitigation technology, and in particular to a quantitative prediction method and system for extended-period precipitation in reservoir areas. Background Technology
[0002] Rainfall is the most significant factor affecting reservoir operation and flood control safety. Accurate long-term rainfall forecasts provide crucial decision-making support for reservoir operation methods and maintenance plans. China experiences the East Asian summer monsoon, followed by a distinct rainy season. The annual flood season sees heavy and concentrated rainfall, posing significant challenges to reservoir power generation and flood control operations. In recent years, frequent extreme rainfall events and unusually high levels of torrential rain have exacerbated the flood control pressure on reservoirs. Furthermore, prolonged full-load operation puts immense pressure on hydropower units, leaving no time for safety maintenance and creating potential safety hazards for both hydropower stations and power generation.
[0003] Currently, reservoir precipitation forecasts generally only predict precipitation within the next 7 days. While rolling forecasts can predict the precipitation process within the next 7 days, they cannot predict the overall precipitation situation for the entire flood season. Precipitation forecasts provided by professional meteorological stations one month in advance are based on climate prediction results and can only predict the probability of more or less precipitation over a large area from the perspective of large-scale circulation. The spatial scope of reservoir basins is relatively small, and the accuracy of forecasts down to a small reservoir area cannot be controlled. This makes it impossible to support the formulation of long-term operation plans and maintenance arrangements for reservoirs. Therefore, it is necessary to conduct research on quantitative forecasting methods for reservoir precipitation over an extended period of 7-30 days to support the safe and economical operation of reservoirs during the flood season and ensure the safety of social power supply. Summary of the Invention
[0004] This invention provides a quantitative prediction method and system for extended-range precipitation in reservoir areas, which solves the technical problem that existing precipitation prediction methods cannot control the accuracy of precipitation prediction when the prediction is refined to a small area of reservoir basin space.
[0005] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows:
[0006] A quantitative prediction method for extended-range precipitation in a reservoir area includes the following steps:
[0007] Collect information on reservoir basins and circulation systems that need to be predicted, use global models to drive mesoscale models, and output daily model circulation factors and model areal rainfall; use a multi-index model system for reservoir basin precipitation prediction to predict daily precipitation forecasts for the reservoir basin over a period of time.
[0008] The global model-driven mesoscale model and reservoir basin precipitation prediction multi-index model system were established through the following steps:
[0009] Collect information on the reservoir basins for which forecasts need to be carried out, including information on the effective runoff generation areas of the reservoirs, reservoir capacity information, and historical precipitation data; collect historical circulation system data, including the position and intensity of the subtropical high, the intensity of the India-Myanmar trough, the upper-level trough, the Northeast cold vortex, and typhoon paths;
[0010] Based on historical precipitation data of the reservoir, precipitation characteristics of the reservoir area were analyzed to obtain a characteristic library of dominant circulation factors;
[0011] Using global model predictions as boundary conditions, a mesoscale model prediction model is driven as a global model-driven mesoscale model.
[0012] Historical precipitation was back-calculated using a global model-driven mesoscale model. The output included large-scale circulation system data and mesoscale grid precipitation data. The large-scale circulation system data included all dominant circulation factors in the dominant circulation factor feature library. The areal rainfall in the reservoir basin was calculated using the mesoscale grid precipitation data and the effective runoff generation area information of the reservoir basin.
[0013] Statistical equations relating model output circulation factor, model output areal rainfall, and actual areal rainfall in the reservoir were established to obtain a multi-index model system for precipitation prediction in the reservoir basin.
[0014] Preferably, the precipitation characteristics analysis of the reservoir area includes the following steps:
[0015] Based on historical precipitation data of the reservoir, identify historical precipitation processes during the flood season in the reservoir basin, and determine the average start and end time of the flood season based on the times of the first and last rounds of heavy precipitation during the flood season.
[0016] Rainfall during the flood season is divided into three categories: heavy rain events, general rainfall events, and prolonged sunny periods.
[0017] Identify three types of precipitation weather events during historical flood seasons, analyze the circulation patterns at that time, and determine the dominant circulation factor characteristic database for heavy rain, general precipitation, and prolonged sunny weather in the reservoir basin.
[0018] Preferably, the reservoir's historical precipitation data, historical circulation system data, historical flood season data, and historical precipitation back-calculation data are all from the past 20 years.
[0019] Preferably, the global model drives the simulation of the physical processes that generate precipitation in the mesoscale model.
[0020] Preferably, a statistical relationship equation is established between the model output circulation factor, the model output areal rainfall, and the actual areal rainfall in the reservoir, using a multiple linear regression method.
[0021] Preferably, the future period is the next 30 days.
[0022] The present invention also provides a computer system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of any of the above methods.
[0023] The present invention has the following beneficial effects:
[0024] The present invention provides a quantitative prediction method and system for extended-range precipitation in reservoir areas, which can achieve mesoscale rainfall prediction in reservoir basins, conduct research on quantitative prediction methods for extended-range reservoir precipitation in the next 7-30 days, support the safe and economical operation of reservoirs during the flood season, and ensure the safety of social power supply.
[0025] In addition to the objectives, features, and advantages described above, the present invention has other objectives, features, and advantages. The invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description
[0026] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:
[0027] Figure 1 This is a flowchart illustrating a preferred embodiment of the present invention for a quantitative prediction method of extended-period precipitation in a reservoir area. Detailed Implementation
[0028] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways as defined and covered by the claims.
[0029] See Figure 1 The quantitative prediction method for extended-range precipitation in reservoir areas according to the present invention includes the following steps:
[0030] S0: Collect information on the reservoir basin and circulation system data that need to be predicted, use the global model to drive the mesoscale model, and output the daily model circulation factors and model areal rainfall; use the multi-index model system for reservoir basin precipitation prediction to predict the daily precipitation forecast results for the reservoir basin in the future.
[0031] The global model-driven mesoscale model and reservoir basin precipitation prediction multi-index model system were established through the following steps:
[0032] S1: Collect information on the reservoir basins for which forecasts need to be carried out, including information on the effective runoff generation areas of the reservoirs, reservoir capacity information, and historical precipitation data; collect historical circulation system data, including the position and intensity of the subtropical high, the intensity of the India-Myanmar trough, the upper-level trough, the Northeast cold vortex, and typhoon paths;
[0033] S2: Based on historical precipitation data of the reservoir, conduct precipitation characteristic analysis of the reservoir area to obtain the characteristic library of dominant circulation factors;
[0034] S3: Using global model prediction results as boundary conditions, drive the mesoscale model prediction model as a global model driving the mesoscale model.
[0035] S4: Historical precipitation is back-calculated using a global model-driven mesoscale model. The output includes large-scale circulation system data and mesoscale grid precipitation data. The large-scale circulation system data includes all dominant circulation factors in the dominant circulation factor feature library. The mesoscale grid precipitation data and the effective runoff generation area information of the reservoir basin are used to calculate the areal rainfall of the reservoir basin.
[0036] Statistical equations relating model output circulation factor, model output areal rainfall, and actual areal rainfall in the reservoir were established to obtain a multi-index model system for precipitation prediction in the reservoir basin.
[0037] Through the above steps, it is possible to predict rainfall in reservoir basins at the mesoscale, conduct research on quantitative prediction methods for reservoir precipitation over an extended period of 7-30 days, support the safe and economical operation of reservoirs during the flood season, and ensure the safety of social power supply.
[0038] The following examples further illustrate this:
[0039] Zhexi Reservoir is a key reservoir in western Hunan Province, with a drainage area of 22,600 square kilometers, a power station with an installed capacity of 447,500 kilowatts, an annual power generation of 2.22 billion kilowatt-hours, and a flood control capacity of 700 million cubic meters. It is an important reservoir for power generation and flood control in western Hunan. As Zhexi Reservoir is a semi-seasonal, semi-annual regulating reservoir, it faces significant flood control pressure and is prone to water wastage during the flood season in high-water years. Therefore, it has a particularly important need for 7-30 day extended-range precipitation forecasting.
[0040] First, a multi-indicator model system for global model-driven mesoscale model and reservoir basin precipitation prediction is established through the following steps:
[0041] S1: Basic Data Collection and Organization
[0042] Information on the Zhexi Reservoir basin was collected, including information on the reservoir's effective runoff generation area, reservoir capacity, and historical day-by-day isal rainfall data for the Zhexi basin from 2000 to 2021. Historical circulation system data from 2000 to 2021 were also collected, including data on the subtropical high's position and intensity, the India-Myanmar trough's intensity, upper-level troughs, the Northeast cold vortex, and typhoon paths.
[0043] S2: Analysis of Precipitation Characteristics in Reservoir Areas
[0044] The average flood season in Hunan Province is from April to August. Based on the daily rainfall data of Zhexi Reservoir from 2000 to 2021, the average start date of the flood season for Zhexi Reservoir is May 19, and the average end date is July 8.
[0045] The precipitation weather during the flood season is divided into three categories: (1) heavy rain process, that is, the daily average rainfall in the reservoir basin exceeds 30 mm and the total average rainfall exceeds 80 mm; (2) general rainfall process, that is, the daily average rainfall in the reservoir basin is 10-20 mm and the total average rainfall is 25-50 mm; (3) continuous sunny process, that is, the daily average rainfall in the reservoir basin is less than 1 mm for 5 consecutive days.
[0046] Identify three types of precipitation weather events during the flood season in the past 20 years, analyze the circulation patterns at that time, and determine the characteristic database of dominant circulation factors for heavy rain, general precipitation, and prolonged sunny weather in the reservoir basin.
[0047] The database of characteristics of rainstorm-dominant circulation factors includes: the subtropical high ridge located at 19°N-20°N, the southern branch trough located at 105°E-108°E, the east-west trending shear line located at 29°N-30°N, etc.
[0048] The dominant factor characteristic database for consecutive sunny days includes: the subtropical high ridge position between 30°N and 32°N, the zonal circulation index >1, etc.
[0049] The dominant factor characteristics of general precipitation weather include: the presence of upper-level troughs and ridges between 100°E and 105°E, and an India-Myanmar trough index > 0, etc.
[0050] S3: Building a monthly precipitation model driven by a global model for mesoscale models.
[0051] The CESM global model and the WRF mesoscale model were selected, with the CESM model having a resolution of 100 km and the WRF model having a resolution of 25 km. A data interface was developed to transfer the CESM global model prediction results to the WRF mesoscale model. Using the CESM model prediction results as boundary conditions, the WRF model was driven to obtain the daily precipitation for the next month.
[0052] S4: Construct a multi-index model system for precipitation prediction in reservoir basins
[0053] A global model was used to drive a mesoscale model to back-calculate historical precipitation events from 2000 to 2021. To conserve computational resources while considering the needs of subsequent operational forecasting, calculations were designed to begin on the first day of each ten-day period from April 1st to August 31st, with a calculation duration of 35 days. The output included large-scale circulation system data and mesoscale gridded precipitation data. The large-scale circulation system data factors included all the dominant circulation factors from step 2. Using mesoscale gridded precipitation data and effective runoff generation area information of the reservoir basin, the areal rainfall of the reservoir basin was calculated. Using a multi-source linear regression method, a statistical relationship equation was established between the model output circulation factors, the model output areal rainfall, and the actual areal rainfall of the reservoir, resulting in a multi-index model system for reservoir basin precipitation forecasting.
[0054] Then make a prediction:
[0055] S0: Conduct 30-day precipitation forecasting and early warning for the reservoir basin.
[0056] As needed, 30-day precipitation forecasts for the reservoir basin are conducted regularly. Global models drive mesoscale models, outputting daily model circulation factors and areal rainfall. Further, the multi-index model system equations for reservoir basin precipitation forecasting are used to obtain daily 30-day precipitation forecasts for the reservoir basin. The forecast results are promptly disseminated to reservoir operation and maintenance and dispatching departments through forecast reports and early warning information.
[0057] In summary, this invention establishes a multi-index model system for predicting reservoir basin precipitation by driving mesoscale models with global models. By extending large-scale global model predictions to mesoscale models, it enables mesoscale rainfall forecasting for reservoir basins. It also facilitates research on quantitative prediction methods for reservoir precipitation over extended periods of 7-30 days, supporting the safe and economical operation of reservoirs during the flood season and ensuring the security of social power supply.
[0058] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for quantitative prediction of extended period precipitation in a reservoir area, characterized by, Includes the following steps: Collect information on reservoir basins and circulation systems that need to be predicted, use global models to drive mesoscale models, and output daily model circulation factors and model areal rainfall; use a multi-index model system for reservoir basin precipitation prediction to predict the daily precipitation forecast results for the next 30 days in the reservoir basin. The global model-driven mesoscale model and the multi-index model system for reservoir basin precipitation prediction are established through the following steps: Collect information on the reservoir basins for which forecasts need to be carried out, including information on the effective runoff generation areas of the reservoirs, reservoir capacity information, and historical precipitation data; collect historical circulation system data, including the position and intensity of the subtropical high, the intensity of the India-Myanmar trough, the upper-level trough, the Northeast cold vortex, and typhoon paths; Based on historical precipitation data of the reservoir, precipitation characteristics of the reservoir area were analyzed to obtain a characteristic library of dominant circulation factors; Using global model predictions as boundary conditions, a mesoscale model prediction model is driven as a global model-driven mesoscale model. Historical precipitation is back-calculated using a global model-driven mesoscale model. The output includes large-scale circulation system data and mesoscale grid precipitation data. The large-scale circulation system data includes all dominant circulation factors in the dominant circulation factor feature library. The areal rainfall in the reservoir basin is calculated using the mesoscale grid precipitation data and the effective runoff generation area information of the reservoir basin. Establish statistical relationship equations between model output circulation factor, model output areal rainfall and actual areal rainfall in reservoir, and obtain a multi-index model system for precipitation prediction in reservoir basin. The analysis of precipitation characteristics in the reservoir area includes the following steps: Based on historical precipitation data of the reservoir, identify historical precipitation processes during the flood season in the reservoir basin, and determine the average start and end time of the flood season based on the times of the first and last rounds of heavy precipitation during the flood season. Rainfall during the flood season is divided into three categories: heavy rain events, general rainfall events, and prolonged sunny periods. Identify three types of precipitation weather events during historical flood seasons, analyze the circulation patterns at that time, and determine the dominant circulation factor characteristic database for heavy rain, general precipitation, and prolonged sunny weather in the reservoir basin.
2. The method for quantitative prediction of reservoir area extended-range precipitation according to claim 1, characterized in that, The historical precipitation data, historical circulation system data, historical flood season data, and historical precipitation calculations for the reservoir are all based on the past 20 years.
3. The method for quantitative prediction of reservoir area extended-range precipitation according to claim 1, characterized in that, The global model drives the simulation of the physical processes that generate precipitation in the mesoscale model.
4. The method for quantitative prediction of reservoir area extended-range precipitation according to claim 1, characterized by, Statistical equations relating model output circulation factor, model output areal rainfall, and actual areal rainfall in the reservoir were established using a multiple linear regression method.
5. A computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.
Citation Information
Patent Citations
Short-time early warning system for severe convection weather
CN105182450A
Monthly runoff prediction method for drainage basin cascade reservoir system
CN113139684A
The system for dynamic downscaling and display climate change senario with a high spatial resoultion
KR1020180079101A
