Meteorological and hydrological forecasting method and system

Abstract

The embodiment of the invention provides a meteorological and hydrological forecasting method and system, and the method comprises the steps: determining a rainfall forecasting data sequence of a target drainage basin according to the climate data of the target drainage basin; inputting the rainfall forecast data sequence and the meteorological and hydrological data of the pre-lag period of the target drainage basin into a target hydrological forecast model for processing to obtain a flow forecast data sequence of the target drainage basin; and superposing the rainfall forecast data sequence and the flow forecast data sequence with the constructed three-dimensional digital model so as to carry out three-dimensional visual display on the forecast data sequence of the target drainage basin through the three-dimensional digital model. The invention aims to improve the accuracy of hydrological forecasting.

Description

Technical Field

[0001] The present application relates to the technical field of meteorological and hydrological forecasting, and in particular to a method and system for meteorological and hydrological forecasting. Background Art

[0002] At present, the daily flow of a river basin is an important part of flood dispatching, the formulation of the operation mode of hydropower station reservoirs, and disaster warning. Accurate prediction of the daily flow of a river basin can achieve efficient use of water resources, thereby significantly improving power generation efficiency and reducing losses caused by floods and droughts. However, due to the many influencing factors involved in the hydrological process within the basin, the traditional hydrological forecasting method does not consider the impact of the basin baseline state, and precipitation forecasts often use a relatively single scheme such as the WRF model for forecasting, resulting in inaccurate forecast results. Summary of the invention

[0003] In view of this, the embodiments of the present application provide a method and system for meteorological and hydrological forecasting, aiming to improve the accuracy of hydrological forecasting.

[0004] A first aspect of an embodiment of the present application provides a method for meteorological and hydrological forecasting, the method comprising:

[0005] Determine the precipitation forecast data sequence of the target basin based on the climate data of the target basin;

[0006] Inputting the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin;

[0007] The precipitation forecast data sequence and the flow forecast data sequence are superimposed on the constructed three-dimensional digital model, so as to perform three-dimensional visualization of the forecast data sequence of the target watershed through the three-dimensional digital model.

[0008] Optionally, based on the climate data of the target basin, a precipitation forecast data sequence for the target basin is determined, including:

[0009] The climate data are simulated through various parameterization schemes of the weather forecast model to determine the initial precipitation forecast data series of the target basin under various parameterization schemes;

[0010] The initial precipitation forecast data series under various parameterized schemes are calculated through the parameterized scheme screening algorithm to determine the representative schemes of the target number;

[0011] The initial precipitation forecast data sequence of the target basin on the current date under each representative scheme is fused through the fusion algorithm to obtain the precipitation forecast data sequence of the target basin.

[0012] Optionally, the initial precipitation forecast data series under various parameterization schemes are calculated through a parameterization scheme screening algorithm to determine the representative schemes of the target number, including:

[0013] According to the initial precipitation forecast data series and the measured precipitation data series of all parameterization schemes, the average precipitation difference data series is determined;

[0014] Determine the parameterization scheme corresponding to the precipitation difference data sequence closest to the average precipitation difference data sequence as the first representative scheme;

[0015] Determine the parameterized scheme corresponding to the precipitation difference data sequence farthest from the precipitation difference data sequence of the first representative scheme as the second representative scheme;

[0016] Calculate the first distance between the precipitation difference data series of the remaining single parameterization schemes and the precipitation difference data series of each representative scheme;

[0017] Determine the smallest first distance among all first distances corresponding to the remaining single parameterized solutions as the target distance of the single parameterized solution;

[0018] According to the target distances of the remaining parameterized solutions, the parameterized solution corresponding to the largest target distance is determined as the next representative solution to obtain a representative solution for the target number.

[0019] Optionally, the initial precipitation forecast data sequence of the target basin on the current date under each representative scheme is fused through a fusion algorithm to obtain the precipitation forecast data sequence of the target basin, including:

[0020] By bringing the initial precipitation forecast data sequence of the target basin on the current date under each representative scheme into the fusion algorithm for calculation, the precipitation forecast data sequence of the target basin is obtained. The expression of the fusion algorithm is:

[0021]

[0022] Among them, p[y(t)|O(t)] represents the precipitation forecast data of the tth day in the precipitation forecast data sequence, p[y(t)|f i (t), O(t)] represents the initial precipitation forecast data of the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, w i (t) represents the weight value corresponding to the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, and n represents the total number of representative schemes.

[0023] Optionally, the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period are input into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin, including:

[0024] According to the pre-hysteresis period of the target meteorological and hydrological data type of the target basin, the meteorological and hydrological data of the target meteorological and hydrological data type within the pre-hysteresis period are input into each target hydrological forecast sub-model in the target hydrological forecast model;

[0025] According to the forecast time corresponding to the precipitation forecast data in the precipitation forecast data sequence, inputting the precipitation forecast data into the target hydrological forecast sub-model corresponding to the forecast time;

[0026] According to the input data, each target hydrological forecast sub-model predicts and obtains the flow forecast data at the corresponding forecast time;

[0027] All flow forecast data obtained by predicting all target hydrological forecast sub-models are determined as the flow forecast data sequence of the target basin.

[0028] Optionally, determine the target meteorological and hydrological data types, including:

[0029] Determine the first month data sequence of the flow of the target basin within the first time period starting from the first moment;

[0030] Determine a second month data sequence of meteorological and hydrological data types starting from a second moment within the first time period, where the second moment is a moment that is separated from the first moment by a preset time period and is before the first moment;

[0031] By performing correlation calculation on the first month's data sequence and the second month's data sequence, a correlation value corresponding to the preset time interval is obtained;

[0032] The comparison result is obtained by comparing the absolute values ​​of the correlation values ​​of the meteorological and hydrological data types at different preset time intervals with the set thresholds respectively;

[0033] When the comparison result indicates that there is an absolute value of a correlation value greater than or equal to a set threshold value among all absolute values ​​of the correlation values, the meteorological and hydrological data type is determined as the target meteorological and hydrological data type.

[0034] Optionally, determine the forward lag period for the target meteorological and hydrological data type, including:

[0035] Determine the maximum absolute value of the correlation value between the flow and the target meteorological and hydrological data type at different preset time intervals;

[0036] The preset time interval corresponding to the absolute value of the maximum correlation value is determined as the front lag period of the target meteorological and hydrological data type.

[0037] Optionally, the method further includes:

[0038] Comparing the flow forecast data series with the flood extreme value to obtain a first comparison result;

[0039] Determine the target forecast time of a flood event according to the first comparison result, and issue a corresponding warning;

[0040] Compare the flow forecast data series with the low water extreme value to obtain a second comparison result;

[0041] According to the second comparison result, the target forecast time of the drought event is determined, and a corresponding warning is issued.

[0042] Optionally, meteorological and hydrological data types include: precipitation, evaporation, flow, and soil moisture.

[0043] Optionally, the target hydrological forecast model is a bidirectional long short-term memory neural network.

[0044] Optionally, determining a target forecast time for a flood event to occur according to the first comparison result, and issuing a corresponding warning, includes:

[0045] When the first comparison result is that there is flow forecast data greater than or equal to the flood extreme value in the flow forecast data sequence, the forecast time corresponding to each flow forecast data greater than or equal to the flood extreme value is determined as the target forecast time, and a corresponding warning is issued;

[0046] Determining the target forecast time of the drought event according to the second comparison result and issuing a corresponding warning includes:

[0047] When the second comparison result is that the flow forecast data of one day in the flow forecast data sequence and the daily flow average of multiple days before that day are less than the low water extreme value, that day is determined as the target forecast time and a corresponding warning is issued.

[0048] A second aspect of the embodiment of the present application provides a system for meteorological and hydrological forecasting, the system comprising a precipitation forecasting module, a hydrological forecasting module and a main body digital module;

[0049] The precipitation forecast module is used to determine the precipitation forecast data sequence of the target basin according to the climate data of the target basin;

[0050] The hydrological forecast module is used to input the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin;

[0051] The main body digital module is used to superimpose the precipitation forecast data sequence and the flow forecast data sequence with the constructed three-dimensional digital model, so as to perform three-dimensional visualization of the forecast data sequence of the target watershed through the three-dimensional digital model.

[0052] Optionally, the main body digital module includes:

[0053] A watershed data module, used for constructing the three-dimensional digital model based on the geographic data of the target watershed;

[0054] The meteorological and hydrological monitoring module is used to monitor and obtain the meteorological and hydrological data of the target basin.

[0055] Optionally, the system further includes a data interaction module and an early warning reporting module;

[0056] The data interaction module is used to store meteorological and hydrological data, and send the predicted precipitation forecast data sequence and flow forecast data sequence to the early warning reporting module for alarm.

[0057] Optionally, the early warning reporting module includes:

[0058] A forecast data reporting module is used to report the predicted precipitation forecast data sequence and flow forecast data sequence;

[0059] The disaster warning module is used to issue a warning when the predicted flow forecast data sequence is analyzed to determine the existence of disaster risks.

[0060] Optionally, the system further comprises an intelligent light storage power supply module, which comprises a solar charging control module, a short message communication module, a storage battery and a controller module;

[0061] The solar charging control module is used to convert the voltage generated by the solar panel into a battery charging voltage;

[0062] SMS communication module, used to receive remote commands sent by the user end at regular intervals and respond;

[0063] Batteries, used for storing and releasing electrical energy;

[0064] The controller module is used to detect the power storage state of the battery according to the instruction information of the user end, and to control the power transmission between the solar charging control module and the battery, and to control the sleep and wake-up of the intelligent power supply.

[0065] The method for meteorological and hydrological forecasting provided by this application has the following advantages:

[0066] A method for meteorological and hydrological forecasting provided in an embodiment of the present application first determines the precipitation forecast data sequence of the target basin according to the climate data of the target basin; inputs the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the previous lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin; superimposes the precipitation forecast data sequence and the flow forecast data sequence with the constructed three-dimensional digital model to perform a three-dimensional visualization display of the forecast data sequence of the target basin through the three-dimensional digital model. Thus, the present application first predicts the precipitation forecast data of the target basin in the next few days to form a precipitation forecast data sequence, and then predicts the flow forecast data of the target basin in the next few days based on the predicted precipitation forecast data sequence and the meteorological and hydrological data of the current past period. In this flow prediction process, the precipitation forecast data and the meteorological and hydrological data of the past period are simultaneously considered, so the accuracy of the hydrological forecast can be effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor.

[0068] Figure 1 A flow chart of a method for meteorological and hydrological forecasting shown as an embodiment of the present application;

[0069] Figure 2 A schematic diagram of a scheme included in a WRF model in a method for meteorological and hydrological forecasting shown as an embodiment of the present application;

[0070] Figure 3 A schematic diagram of a meteorological and hydrological forecasting system is shown as an embodiment of the present application. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0072] Figure 1The following is a flow chart of a method for meteorological and hydrological forecasting according to an embodiment of the present application. Figure 1 As shown, the method includes:

[0073] Step S1: Determine the precipitation forecast data sequence of the target basin based on the climate data of the target basin.

[0074] In this embodiment, the climate data of the target basin is obtained, and the climate data of the target basin is simulated by a weather forecast model to determine the precipitation forecast data sequence of the target basin, and the precipitation forecast data sequence is the precipitation forecast data of each day from the first day to the Nth day in the future, and the N value is preferably 15 days. The target basin can be any river basin.

[0075] In this embodiment, the climate data of the target basin can be obtained through the ERA5 data set, which is the fifth generation ECMWF atmospheric reanalysis of the global climate from January 1950 to the present by ECMWF (European Center for Medium-Range Weather Forecasts). ERA5 is produced by the Copernicus Climate Change Service (C3S) of ECMWF, providing hourly estimates of a large number of atmospheric, land and ocean climate variables. ERA5 combines model data with observational data from all over the world to form a global complete and consistent data set. The weather forecast model is preferably the WRF model (Weather Research and Forecasting Model).

[0076] Step S2: Input the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin.

[0077] In this embodiment, after determining the precipitation forecast data sequence after the current moment of the target basin through step S1, the precipitation forecast data sequence and the meteorological and hydrological data in the pre-hysteresis period before the current moment of the target basin are input into the pre-trained target hydrological forecast model for processing, and a flow forecast data sequence with the same sequence length as the precipitation forecast data sequence is obtained. The pre-hysteresis of the target basin refers to how long before the current moment the meteorological and hydrological data will affect the future flow forecast results. For example, the meteorological and hydrological data within 2 months before the current moment of the target basin will affect the future flow forecast results, so the 2 months before the current moment are determined as the pre-hysteresis of the target basin. Among them, the meteorological and hydrological data include but are not limited to precipitation, evaporation, flow, and soil moisture. For the pre-hysteresis of meteorological and hydrological data, a corresponding pre-hysteresis can be determined for each different type of meteorological and hydrological data. For example, for the meteorological and hydrological data of precipitation, the corresponding pre-hysteresis is determined to be 2 months, and for the meteorological and hydrological data of evaporation, the corresponding pre-hysteresis is determined to be 3 months, etc.

[0078] In the present embodiment, Long Short-Term Memory (LSTM) is a deep learning neural network that can effectively handle long time series problems. It uses memory cells to replace the cell units in a simple recurrent neural network (RNN), thereby solving the gradient vanishing or gradient explosion problems that RNN may encounter during training. The LSTM model consists of an input layer memory cell and an output layer, wherein the memory cell has the ability to selectively memorize or forget information. However, traditional unidirectional long and short-term memory can only be encoded unidirectionally, so it can only process information from front to back and cannot process information from back to front, so it may not be able to achieve the best results when processing time series problems with strong front and back correlation. In order to solve this problem, a bidirectional long and short-term memory neural network (BiLSTM, Bi-directional Long Short-Term Memory) can effectively capture the correlation between data when processing time series data by using both forward and reverse neural networks, so BiLSTM is more suitable than LSTM for hydrological forecasting problems with dependencies between front and back data, and is easier to capture the characteristics of hydrological time series changes. The BiLSTM model predicts based on the entire time series data by fusing forward and reverse LSTM neural networks. The forward LSTM layer starts inputting from the start of the sequence until it stops at the end, and the reverse LSTM layer reads in reverse. The two LSTM layers are finally connected to the same output layer. In this way, the BiLSTM model can fuse the past information features extracted by the forward LSTM network and the future information features extracted by the reverse LSTM network, thereby enhancing the expression ability and generalization ability of the model. Therefore, the target hydrological forecast model in this application is preferably a BiLSTM model. It should be understood that this is only a preferred implementation of the target hydrological forecast model, which can also be other deep learning models, and is not specifically limited here. The present application preferably selects the hyperparameters of the model through the gray wolf optimization algorithm. The advantage of using this optimization algorithm is that it automatically searches the parameter space and finds the optimal parameter combination, which not only improves the prediction performance, but also saves a lot of time and energy. Changing a single hyperparameter value cannot achieve the purpose of model optimization, but a better hyperparameter combination can be easily obtained through the gray wolf algorithm. Different hyperparameters will play different roles in the model. For example, the learning rate determines the speed at which the parameters are updated in each iteration in the deep learning model. However, if the learning rate is too large, the model may not converge, and if the learning rate is too small, the model may fall into local extreme values. The number of units in deep learning models refers to the number of hidden layer units. The number of units can make the model fit the data better. However, a large number of units may cause the model to overfit, and a small number of units may reduce the complexity of the model and also cause the model to underfit. In addition, the L2 regularization parameter is a penalty term that controls the complexity of the model.A larger L2 regularization parameter value will reduce the complexity of the model, but it will also increase the risk of underfitting the model. In the case of coupled data, the model hyperparameters need to be selected and adjusted one by one. The gray wolf algorithm can be used to select the model hyperparameters one by one, which does not rely on manual experience, reduces the difficulty of use, and improves the prediction accuracy of the model.

[0079] Step S3: superimposing the precipitation forecast data sequence and the flow forecast data sequence with the constructed three-dimensional digital model, so as to perform three-dimensional visualization of the forecast data sequence of the target watershed through the three-dimensional digital model.

[0080] In this embodiment, after the precipitation forecast data sequence and the flow forecast data sequence of the target river basin are predicted and obtained through step S1 and step S2, the precipitation forecast data sequence and the flow forecast data sequence are superimposed on a pre-constructed three-dimensional digital model corresponding to a larger river basin including the target river basin, so as to visualize the precipitation forecast data sequence and the flow forecast data sequence of the target river basin through the three-dimensional digital model.

[0081] In this embodiment, the construction of the three-dimensional digital model specifically includes: within the determined watershed, using satellite remote sensing images, river system data, 30m resolution digital elevation data (DEM, Digital Elevation Model), and water conservancy project distribution data to construct a three-dimensional digital model. At the same time, for the key river sections within the watershed, using drone oblique photogrammetry and BIM three-dimensional modeling, an OSGB format oblique photography three-dimensional model is generated, and the terrain results of digital orthophoto (DOM, Digital Orthophoto Map), DEM, and digital surface model (DSM, Digital Surface Model) data are simultaneously provided, and the key river sections in the constructed three-dimensional digital model are refined. Model construction is carried out to obtain the final three-dimensional digital model of the watershed, and the watershed range corresponding to the three-dimensional digital model includes the target watershed. At the same time, on the constructed three-dimensional digital model, the dynamic and static data of the river network can also be superimposed with all elements to construct a twin panorama of the river network for dynamic interaction and real-time fusion simulation of data such as flow and precipitation, and realize the three-dimensional visualization evolution of the water level, flow direction, flow velocity, flow, and precipitation of the river network.

[0082] A method for meteorological and hydrological forecasting provided in an embodiment of the present application first determines the precipitation forecast data sequence of the target basin according to the climate data of the target basin; inputs the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the previous lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin; superimposes the precipitation forecast data sequence and the flow forecast data sequence with the constructed three-dimensional digital model to perform a three-dimensional visualization display of the forecast data sequence of the target basin through the three-dimensional digital model. Thus, the present application first predicts the precipitation forecast data of the target basin in the next few days to form a precipitation forecast data sequence, and then predicts the flow forecast data of the target basin in the next few days based on the predicted precipitation forecast data sequence and the meteorological and hydrological data of the current past period. In this flow prediction process, the precipitation forecast data and the meteorological and hydrological data of the past period are simultaneously considered, so the accuracy of the hydrological forecast can be effectively improved.

[0083] In combination with the above embodiments, in one implementation, the present application embodiment further provides a method for meteorological and hydrological forecasting. In the method for meteorological and hydrological forecasting, step S1 may include steps S11 to S13:

[0084] Step S11: Simulate the climate data through various parameterization schemes of the weather forecast model to determine the initial precipitation forecast data sequence of the target basin under various parameterization schemes.

[0085] In this embodiment, the current climate data of the target basin is simulated through multiple parameterization schemes of the weather forecast model to obtain the initial precipitation forecast data sequence under the target basin corresponding to the parameterization scheme, and each parameterization scheme has a corresponding initial precipitation forecast data sequence. Figure 2 As shown, the multiple parameterization schemes at least include the microphysical process scheme, land surface process scheme, planetary boundary layer scheme and cumulus parameterization scheme in the WRF model, and each type of parameterization scheme includes multiple sub-schemes. When the multiple parameterization schemes include the microphysical process scheme, the land surface process scheme, the planetary boundary layer scheme and the cumulus parameterization scheme, the microphysical process scheme in the multiple parameterization schemes can be any sub-scheme in the microphysical process scheme, the land surface process scheme in the multiple parameterization schemes can be any sub-scheme in the land surface process scheme, the planetary boundary layer scheme in the multiple parameterization schemes can be any sub-scheme in the planetary boundary layer scheme, and the cumulus parameterization scheme in the multiple parameterization schemes can be any sub-scheme in the cumulus parameterization scheme. The sequence length of the initial precipitation forecast data sequence is preferably 15, that is, a sequence composed of the initial precipitation forecast data for the next 15 days.

[0086] Step S12: Calculate the initial precipitation forecast data sequence under various parameterized schemes through the parameterized scheme screening algorithm to determine the representative schemes of the target number.

[0087] In this embodiment, the initial precipitation forecast data sequence corresponding to the various parameterized schemes determined in step S11 is calculated by the parameterized scheme screening algorithm to determine the parameterized scheme with the target number that performs better, and the parameterized scheme with the target number is the representative scheme of the determined target number. The target number can be set according to the actual application scenario, and is not specifically limited here, such as 3, 5, etc., but the value will be less than the total number of all parameterized schemes.

[0088] In the present application, step S12 may include steps S121 to S126:

[0089] Step S121: Determine the average precipitation difference data sequence according to the initial precipitation forecast data sequence and the measured precipitation data sequence of all parameterized schemes.

[0090] In this embodiment, the initial precipitation forecast data sequence and the measured precipitation data sequence of all parameterized schemes obtained by simulation are calculated by the first algorithm to obtain a corresponding average precipitation difference data sequence, wherein the expression of the first algorithm is:

[0091]

[0092] Where n represents the nth parameterization scheme, and its value is [1, N]. N represents the total number of all parameterization schemes obtained by simulation. n-m It is the difference between the initial precipitation forecast data for the mth day in the initial precipitation forecast data sequence of the nth parameterization scheme and the actual precipitation data for the mth day in the actual precipitation data sequence.

[0093] In this embodiment, there is a unified measured precipitation data sequence for the initial precipitation forecast data sequence of each parameterized scheme, and the initial precipitation forecast data sequence records the predicted initial precipitation forecast data based on the specified date for the next 15 days, while the corresponding measured precipitation data sequence records the actual measured precipitation data based on the specified date for the next 15 days. For example, the 16th day from the current time is regarded as a specific day, and the past 15 days from the current time belong to the future 15 days of the specific day, and the initial precipitation forecast data of the future 15 days can be obtained by prediction, and at the same time, the future 15 days for the current time that has already occurred can obtain the measured precipitation data of the future 15 days.

[0094] In this embodiment, the precipitation difference data sequence corresponding to the parameterized scheme is obtained by subtracting the initial precipitation forecast data sequence of the parameterized scheme from the measured precipitation data sequence. The i-th data in the precipitation difference data sequence represents the difference between the initial precipitation forecast data of the future i-th day in the initial precipitation forecast data sequence of the parameterized scheme and the measured precipitation data of the future i-th day in the measured precipitation data sequence.

[0095] Step S122: Determine the parameterization scheme corresponding to the precipitation difference data sequence closest to the average precipitation difference data sequence as the first representative scheme.

[0096] In this embodiment, after the average precipitation difference data sequence is determined in step S122, the distances between the precipitation difference data sequences of each parameterized scheme and the average precipitation difference data sequence are calculated by a distance algorithm, and then the parameterized scheme corresponding to the precipitation difference data sequence closest to the average precipitation difference data sequence is determined as the first representative scheme.

[0097] The distance algorithm is preferably a Euclidean distance algorithm. It should be understood that this is only a preferred implementation of the distance algorithm. The distance algorithm may also be other distance algorithms, which are not specifically limited here.

[0098] In this embodiment, when the distance algorithm is the Euclidean distance algorithm, the calculation expression is:

[0099]

[0100] In this calculation expression, y i represents the average precipitation difference data for the future i-th day in the average precipitation difference data sequence, x i It represents the precipitation difference data of the future i-th day in the precipitation difference data sequence of any parameterization scheme among all parameterization schemes obtained by simulation.

[0101] Step S123: Determine the parameterized scheme corresponding to the precipitation difference data sequence farthest from the precipitation difference data sequence of the first representative scheme as the second representative scheme.

[0102] In this embodiment, after the first representative scheme is determined in step S122, the distance between the precipitation difference data sequences of all other parameterized schemes except the first representative scheme and the precipitation difference data sequence of the first representative scheme is calculated by a distance algorithm, and then the parameterized scheme corresponding to the precipitation difference data sequence farthest from the precipitation difference data sequence of the first representative scheme is determined as the second representative scheme. If the target number is 2, then the first representative scheme and the second representative scheme can be determined as the representative schemes of the target number, and if the target number is greater than 2, step S124 is executed to continue determining the representative scheme.

[0103] Step S124: Calculate the first distance between the precipitation difference data series of the remaining single parameterized scheme and the precipitation difference data series of each representative scheme.

[0104] In this embodiment, the first distances between the precipitation difference data sequences of all other parameterized schemes except the one that has been determined as the representative scheme and the precipitation difference data sequences of each representative scheme determined as the representative scheme are calculated by a distance algorithm. Among them, whether it is the first representative scheme, the second representative scheme or the subsequent representative scheme determined by step S124, they all belong to the determined representative schemes.

[0105] For example, when representative schemes a1, a2, and a3 have been determined, and the remaining parameterized schemes that have not been determined as representative schemes are parameterized schemes a4, a5, and a6, a first distance a41 between the precipitation difference data sequence of parameterized scheme a4 and the precipitation difference data sequence of representative scheme a1 is calculated by a distance algorithm, a first distance a42 between the precipitation difference data sequence of parameterized scheme a4 and the precipitation difference data sequence of representative scheme a2 is calculated by a distance algorithm, and a first distance a43 between the precipitation difference data sequence of parameterized scheme a4 and the precipitation difference data sequence of representative scheme a3 is calculated by a distance algorithm. The first distance a43, corresponding to the parameterized scheme a4 that has not been determined as a representative scheme, will obtain three first distances with three different representative schemes, which are a41, a42, and a43 respectively; based on the same calculation method, the parameterized scheme a5 that has not been determined as a representative scheme will obtain three first distances with three different representative schemes, which are a51, a52, and a53 respectively; based on the same calculation method, the parameterized scheme a6 that has not been determined as a representative scheme will obtain three first distances with three different representative schemes, which are a61, a62, and a63 respectively.

[0106] Step S125: Determine the smallest first distance among all first distances corresponding to the remaining single parameterization schemes as the target distance of the single parameterization scheme.

[0107] In this embodiment, for multiple first distances between any parameterized scheme that has not been determined as a representative scheme and multiple representative schemes, a first distance with the smallest value among the multiple first distances is selected to be determined as the target distance of any parameterized scheme. Based on the same implementation method, each parameterized scheme that has not been determined as a representative scheme will obtain a target distance corresponding to itself.

[0108] For example, continuing to use the example in the above step S124, for the three first distances a41, a42, and a43 between the parameterized scheme a4 and the three different representative schemes, the value of a42 is the smallest, so a42 is determined as the target distance of the parameterized scheme a4; for the three first distances a51, a52, and a53 between the parameterized scheme a5 and the three different representative schemes, the value of a51 is the smallest, so a51 is determined as the target distance of the parameterized scheme a5; for the three first distances a61, a62, and a63 between the parameterized scheme a6 and the three different representative schemes, the value of a62 is the smallest, so a62 is determined as the target distance of the parameterized scheme a6.

[0109] Step S126: According to the target distances of the remaining parameterized solutions, determine the parameterized solution corresponding to the largest target distance as the next representative solution, so as to obtain representative solutions of the target number.

[0110] In this embodiment, after determining the target distances of all remaining parameterized solutions that have not been determined as representative solutions, the parameterized solution corresponding to the largest target distance is determined as the representative solution. At this time, it is determined whether the number of representative solutions obtained has reached the target number. If it has not reached the target number, steps S124 to S126 are continuously executed to determine new representative solutions until the target number of representative solutions is obtained.

[0111] For example, the example in step S125 is continued to be used, and the target distance a42 of parameterized solution a4, the target distance a51 of parameterized solution a5, and the target distance a62 of parameterized solution a6 are determined. Then, it is determined that a51 is the largest among the target distances a42, a51, and a62, and thus the parameterized solution a5 corresponding to the target distance a51 is determined as the representative solution. If the four representative solutions a1, a2, a3, and a5 determined at this time do not reach the target number, then the first distances between the remaining parameterized solutions a4 and a6 and the representative solutions a1, a2, a3, and a5 are determined, respectively, and then the smallest first distance among the multiple first distances corresponding to the parameterized solution itself is determined as the target distance of the parameterized solution. After obtaining the target distance of parameterized solution a4 and the target distance of parameterized solution a6, the parameterized solution corresponding to the largest target distance of the two target distances is determined as the representative solution, and this cycle is repeated until the representative solution of the target number is obtained.

[0112] Step S13: The initial precipitation forecast data sequence of the target basin on the current date under each representative scheme is fused through a fusion algorithm to obtain the precipitation forecast data sequence of the target basin.

[0113] In this embodiment, after the representative schemes of the target number are determined through steps S121 to S126, the initial precipitation forecast data sequences under various representative parameter schemes are determined starting from the current date. For example, the representative schemes include scheme 1, scheme 2 and scheme 3. If the current date is 202x-y-month-z-day, then from the z-day, the precipitation forecast data for the next 15 days is predicted by scheme 1 to obtain the initial precipitation forecast data sequence corresponding to scheme 1. From the z-day, the precipitation forecast data for the next 15 days is predicted by scheme 2 to obtain the initial precipitation forecast data sequence corresponding to scheme 2. From the z-day, the precipitation forecast data for the next 15 days is predicted by scheme 3 to obtain the initial precipitation forecast data sequence corresponding to scheme 3. Then, the initial precipitation forecast data sequences of the target basin on the current date under each representative scheme are merged to obtain a final precipitation forecast data sequence for the target basin.

[0114] In the present application, step S13 may include: obtaining the precipitation forecast data sequence of the target basin by bringing the initial precipitation forecast data sequence of the current date of the target basin under each representative scheme into the fusion algorithm for calculation, and the expression of the fusion algorithm is:

[0115]

[0116] Among them, p[y(t)|O(t)] represents the precipitation forecast data of the tth day in the precipitation forecast data sequence, p[y(t)|f i(t), O(t)] represents the initial precipitation forecast data of the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, w i (t) represents the weight value corresponding to the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, and n represents the total number of representative schemes.

[0117] In this embodiment, after the representative schemes of the target quantity are determined through steps S121 to S126, the initial precipitation forecast data sequences of the target basin on the current date under each representative scheme are fused through a fusion algorithm to obtain a final precipitation forecast data sequence of the target basin. The expression of the fusion algorithm is:

[0118]

[0119] Among them, p[y(t)|O(t)] represents the precipitation forecast data for the next t day in the precipitation forecast data sequence, and p[y(t)|f i (t), O(t)] represents the initial precipitation forecast data for the tth day in the initial precipitation forecast data sequence under the ith representative scheme, w i (t) represents the weight value corresponding to the future t-th day in the initial precipitation forecast data sequence under the i-th representative scheme, and n represents the total number of representative schemes.

[0120] In combination with the above embodiments, in one implementation, the present application embodiment further provides a method for meteorological and hydrological forecasting. In the method for meteorological and hydrological forecasting, step S2 may include steps S21 to S24:

[0121] Step S21: according to the pre-hysteresis period of the target meteorological and hydrological data type of the target basin, the meteorological and hydrological data of the target meteorological and hydrological data type within the pre-hysteresis period are input into each target hydrological forecast sub-model in the target hydrological forecast model.

[0122] In this embodiment, the target hydrological forecast model includes multiple target hydrological forecast sub-models, and the number of the multiple target hydrological forecast sub-models is the same as the sequence length of the precipitation forecast data sequence. For example, when the precipitation forecast data sequence obtained by simulation is the precipitation forecast data for each day in the next 15 days, the number of the target hydrological forecast sub-models is also 15.

[0123] In this embodiment, for each target hydrological forecast sub-model in the target hydrological forecast model, the input data is the same, that is, the meteorological and hydrological data within the pre-lag period of the target meteorological and hydrological data type of the target basin is input into each target hydrological forecast sub-model.

[0124] For example, when the target meteorological and hydrological data types include evaporation type, flow type and soil moisture type, and the lead lag period of the evaporation type is 2 months, the lead lag period of the flow type is 1 month, and the lead lag period of the soil moisture type is 3 months, the evaporation data of the evaporation type 2 months before the current moment is input into each target hydrological forecast sub-model, and the flow data of the flow type 1 month before the current moment is input into each target hydrological forecast sub-model, and the soil moisture data of the soil moisture type 3 months before the current moment is input into each target hydrological forecast sub-model.

[0125] Step S22: According to the forecast time corresponding to the precipitation forecast data in the precipitation forecast data sequence, the precipitation forecast data is input into the target hydrological forecast sub-model corresponding to the forecast time.

[0126] In this embodiment, any one of the target hydrological forecast sub-models in the target hydrological forecast model is used to predict the flow rate for the next day corresponding to itself. The data input to any one of the target hydrological forecast sub-models is the precipitation forecast data corresponding to the next day in the precipitation forecast data sequence obtained by simulation. That is, each target hydrological forecast sub-model inputs the precipitation forecast data for the day corresponding to its own forecast time.

[0127] For example, the sequence length of the final precipitation forecast data sequence obtained by simulation is 15 days, which corresponds to the precipitation forecast data for the next 15 days. There are 15 target hydrological forecast sub-models m1 to m15, and the i-th target hydrological forecast sub-model mi among the 15 target hydrological forecast sub-models is used to predict the flow for the future i-th day, and i takes a value of 1 to 15. Correspondingly, the precipitation forecast data for the future i-th day is input into the i-th target hydrological forecast sub-model mi.

[0128] Step S23: Based on the input data, each target hydrological forecast sub-model predicts and obtains the flow forecast data at the corresponding forecast time.

[0129] In this embodiment, each target hydrological forecast sub-model predicts the flow rate for the next day corresponding to itself based on the input data inputted into itself, and obtains a corresponding flow rate forecast data.

[0130] Step S24: All flow forecast data obtained by predicting all target hydrological forecast sub-models are determined as the flow forecast data sequence of the target river basin.

[0131] In this embodiment, after obtaining the flow forecast data predicted by all target hydrological forecast sub-models, all the predicted flow forecast data are sorted in chronological order to obtain a flow forecast data sequence for the target basin.

[0132] In combination with the above embodiments, in one implementation, the present application embodiment further provides a method for meteorological and hydrological forecasting. In the method for meteorological and hydrological forecasting, the target meteorological and hydrological data type is determined, including steps S01 to S05:

[0133] Step S01: Determine the first month data sequence of the flow of the target basin within the first time period starting from the first moment.

[0134] In this embodiment, before the flow prediction is performed, the target meteorological and hydrological data types involved in the target hydrological forecast model processing are determined, that is, which meteorological and hydrological data types will have a greater impact on the flow in the future. The first month data series refers to the monthly data series of the flow.

[0135] Specifically, the first moment is a moment before the flow prediction begins. At this time, the monthly data sequence of the flow of the target basin in the past first time period is determined from the first moment, and the monthly data sequence is determined as the first monthly data sequence. The first time period can be set according to the actual application scenario, and can be the past 10 years, 5 years, etc. before the first moment, and is not specifically limited here.

[0136] Step S02: Determine a second month data sequence of meteorological and hydrological data types starting from a second moment within a first time period, wherein the second moment is a moment that is separated from the first moment by a preset time period and is before the first moment.

[0137] In this embodiment, for a type of meteorological and hydrological data, a monthly data sequence of the meteorological and hydrological data type within a past first time period is determined starting from a second moment, and the monthly data sequence is determined as the second monthly data sequence of the meteorological and hydrological data type, wherein the second moment is a moment that is separated from the first moment by a preset time interval and is before the first moment.

[0138] For example, for the meteorological and hydrological data type of evaporation, when the first time period is the past 10 years from the first moment, the preset time interval is 1 month, and the first moment is December 2020 in the past, the first month data series of flow is the monthly flow series of the past 10 years from January 2011 to December 2020. The second month data series of evaporation type is the monthly evaporation series of the past 10 years from December 2010 to November 2020.

[0139] Step S03: performing correlation calculation on the first month's data sequence and the second month's data sequence to obtain a correlation value corresponding to the preset time interval.

[0140] In this embodiment, after determining the first month data sequence of the flow of the target basin and the second month data sequence of a type of meteorological and hydrological data through step S01 and step S02, the first month data sequence and the second month data sequence are correlated to obtain a correlation value corresponding to a preset time interval. The expression for correlation calculation is:

[0141]

[0142] Among them, x i is the monthly data of the ith month in the first month data sequence, y i is the monthly data of the ith month in the second month data sequence, is the monthly average data of the first month’s data series, is the monthly average data of the second month's data sequence, n is the sequence length of the monthly data sequence, and the closer the correlation coefficient value is to -1 or 1, the stronger the correlation between the two variables is.

[0143] For example, continuing with the example in the above step S03, the monthly flow sequence for the past 10 years from January 2011 to December 2020 (the first month data sequence belonging to the flow) and the monthly evaporation sequence for the past 10 years from December 2010 to November 2020 (the second month data sequence belonging to the evaporation type) are correlated through the above correlation calculation expression to obtain the corresponding correlation value, which belongs to the correlation value corresponding to the time interval of 1 month.

[0144] Step S04: Obtaining comparison results by comparing the absolute values ​​of the correlation values ​​of the meteorological and hydrological data types at different preset time intervals with set thresholds.

[0145] In this embodiment, for a type of meteorological and hydrological data, the correlation values ​​at multiple different preset time intervals are calculated by the same implementation as above, and then the absolute values ​​of the correlation values ​​at multiple different preset time intervals are compared with the set threshold value to obtain the corresponding comparison result. For example, the correlation value between a type of meteorological and hydrological data and the flow rate when the time interval is 1 month, the correlation value between the type of meteorological and hydrological data and the flow rate when the time interval is 2 months, the correlation value between the type of meteorological and hydrological data and the flow rate when the time interval is 3 months, and the correlation value between the type of meteorological and hydrological data and the flow rate when the time interval is 4 months can be calculated.

[0146] Step S05: when the comparison result represents that there is an absolute value of a correlation value greater than or equal to a set threshold among all the absolute values ​​of the correlation values, the meteorological and hydrological data type is determined as the target meteorological and hydrological data type.

[0147] In this embodiment, for the comparison result obtained by step S04, when the absolute value of the correlation value at all different preset time intervals is less than the set threshold, it is determined that the meteorological and hydrological data type has a low correlation with the flow of the target basin, so it is not determined as the target meteorological and hydrological data type, and the meteorological and hydrological data type will not participate in the subsequent flow prediction process through the target hydrological forecast model. In the case that there is an absolute value of the correlation value greater than or equal to the set threshold among the absolute values ​​of the correlation values ​​at all different preset time intervals, it is determined that the meteorological and hydrological data type has a high correlation with the flow of the target basin, so it is determined as the target meteorological and hydrological data type, and the target meteorological and hydrological data type will participate in the subsequent flow prediction process through the target hydrological forecast model. Among them, the set threshold can be set according to the actual application scenario, and is not specifically limited here. Therefore, for various meteorological and hydrological data types, it can be determined whether it will be determined as the target meteorological and hydrological data type participating in the subsequent flow prediction through the same implementation method of steps S01 to S05, wherein the various meteorological and hydrological data types here include but are not limited to precipitation, evaporation, flow, and soil moisture.

[0148] In combination with the above embodiments, in one implementation, the embodiment of the present application further provides a method for meteorological and hydrological forecasting. In the method for meteorological and hydrological forecasting, determining the front lag period of the target meteorological and hydrological data type includes: determining the maximum absolute value of the correlation value among the absolute values ​​of the correlation values ​​between the flow and the target meteorological and hydrological data type at different preset time intervals; and determining the preset time interval corresponding to the maximum absolute value of the correlation value as the front lag period of the target meteorological and hydrological data type.

[0149] In this embodiment, after each target meteorological and hydrological data type is determined in step S05, for any target meteorological and hydrological data type, the size of the correlation value corresponding to each different preset time interval is determined, and the preset time interval corresponding to the largest absolute value of the correlation value is determined as the pre-hysteresis period of the target meteorological and hydrological data type. For example, for the target meteorological and hydrological data type b1, it includes the correlation value x1 corresponding to the preset time interval of 1 month, the correlation value x2 corresponding to the preset time interval of 2 months, the correlation value x3 corresponding to the preset time interval of 3 months, and the correlation value x4 corresponding to the preset time interval of 4 months, wherein the absolute value |x3| of the correlation value corresponding to the preset time interval of 3 months is the largest, so 3 months is determined as the pre-hysteresis period of the target meteorological and hydrological data type b1. Therefore, when performing flow prediction, the data of the target meteorological and hydrological data type b1 input into the target hydrological forecast model is the data of the past 3 months.

[0150] In combination with the above embodiments, in one implementation, the present application embodiment further provides a method for meteorological and hydrological forecasting. In the method for meteorological and hydrological forecasting, the method further includes steps S4 to S7:

[0151] Step S4: Compare the flow forecast data sequence with the flood extreme value to obtain a first comparison result.

[0152] In this embodiment, after the flow forecast data sequence is obtained, the flow forecast data sequence is compared with the flood extreme value to obtain a corresponding first comparison result.

[0153] Step S5: According to the first comparison result, determine the target forecast time of the flood event and issue a corresponding warning.

[0154] In this embodiment, when the first comparison result is that there is flow forecast data greater than or equal to the flood extreme value in the flow forecast data sequence, the forecast time corresponding to each flow forecast data greater than or equal to the flood extreme value is determined as the time when a flood event may occur, and the corresponding forecast time is the target forecast time, and then early warnings are issued for these future target forecast times. For example, the flow forecast data sequence includes flow forecast data for the next 15 days, where the flow forecast data for the next 3rd and 5th days is greater than the flood extreme value, so it is determined that a flood event may occur on the next 3rd and next 5th days, so corresponding early warnings are issued for these two days. This application determines the 95% quantile of the daily flow sequence within a period of history as the extreme high runoff (Q95) for that period, and determines the extreme high runoff as the flood extreme value.

[0155] Step S6: Compare the flow forecast data sequence with the low water extreme value to obtain a second comparison result.

[0156] In this embodiment, after the flow forecast data sequence is obtained, the flow forecast data sequence is compared with the low water extreme value to obtain a corresponding second comparison result.

[0157] Step S7: According to the second comparison result, determine the target forecast time of the drought event and issue a corresponding warning.

[0158] In this embodiment, when the second comparison result is that the flow forecast data of one day in the flow forecast data sequence and the daily flow average of 1 to 6 days before that day (the daily flow average can be the forecast flow, and can also be the measured flow. If there is a measured flow, the measured flow is used first) is less than the low water extreme value, it is determined that a drought event may occur on that day, and then a corresponding warning is issued. This application determines the 5% quantile of the flow average sequence of 7 consecutive days in a historical period as an extremely low runoff (Q5), and determines the extremely low runoff as the low water extreme value.

[0159] In this embodiment, when using the deep learning model for hydrological forecasting, a target hydrological forecast sub-model is established for each future day, because different dates in the future have different degrees of dependence on the historical state of the basin. This method can improve the accuracy of hydrological forecasting and promote the efficient use of water resources in cascade reservoirs. When forecasting the flow in the next 1-15 days, the meteorological conditions in the next 1-15 days and the meteorological and hydrological data types such as precipitation, flow, evaporation, and soil moisture in the past specified months are considered at the same time to improve the scientific nature of the prediction results. This is because the flow generation of the basin is closely related to the state of the basin. Based on the ERA5 data-driven WRF climate model, multiple representative schemes are selected based on the representative scheme selection method, and the selected multiple representative schemes are integrated to improve the resolution and accuracy of meteorological forecast products. Other effects are the realization of visual presentation, IoT perception control, digital expression, and data fusion supply. In terms of visual presentation: physical entities and business processes are displayed through real-time rendering, and the overall scene of the basin is displayed until the details of the project, providing visual and intuitive effects, such as precipitation, flood and other scene rendering. In terms of IoT perception and control: it mainly perceives various elements and remotely controls various control devices to provide data sources for digital three-dimensional models. On the basis of the traditional water conservancy monitoring system, intelligent perception technology and communication technology are used to build an integrated sky-ground water conservancy perception network that monitors and intelligently analyzes the properties of water-related objects and their environmental status. The prediction model is combined to realize meteorological data prediction, and then combined with the deep learning model to realize hydrological forecast. In terms of digital expression: it mainly constructs a three-dimensional digital model to organize data of different types, forms, and sources generated by the physical watershed in an orderly manner. In terms of data fusion supply: it provides massive data services for models and business applications in the form of data streams.

[0160] Based on the same inventive concept, an embodiment of the present application provides a system for meteorological and hydrological forecasting, such as Figure 3 As shown, the system 300 includes a precipitation forecast module 301, a hydrological forecast module 302 and an entity digital module 303;

[0161] The precipitation forecast module 301 is used to determine the precipitation forecast data sequence of the target basin according to the climate data of the target basin;

[0162] The hydrological forecast module 302 is used to input the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin;

[0163] The main body digital module 303 is used to superimpose the precipitation forecast data sequence and the flow forecast data sequence with the constructed three-dimensional digital model, so as to perform three-dimensional visualization of the forecast data sequence of the target watershed through the three-dimensional digital model.

[0164] In this embodiment, the hydrological forecast module 302 standardizes and packages the trained target hydrological forecast model to construct a model platform, and provides model services with standardized interfaces for business applications.

[0165] Optionally, the main body digital module 303 includes:

[0166] A watershed data module, used for constructing the three-dimensional digital model based on the geographic data of the target watershed;

[0167] The meteorological and hydrological monitoring module is used to monitor and obtain the meteorological and hydrological data of the target basin.

[0168] In this embodiment, by deploying multiple meteorological monitoring sensors in the river basin, real-time monitoring, transmission and storage of variables such as wind speed, temperature, humidity, precipitation, etc. are realized, and the meteorological and hydrological monitoring module is used to store the daily meteorological data of the river basin in the past specified time period in real time; at the same time, multiple hydrological monitoring sensors are deployed in the river basin to realize real-time monitoring, transmission and storage of variables such as flow, soil moisture content, evaporation, water level, etc., and the meteorological and hydrological monitoring module is used to store the daily hydrological data of the river basin in the past specified time period in real time.

[0169] Optionally, the system further includes a data interaction module and an early warning reporting module;

[0170] The data interaction module is used to store meteorological and hydrological data, and send the predicted precipitation forecast data sequence and flow forecast data sequence to the early warning reporting module for alarm.

[0171] In this embodiment, the data interaction module is used to receive, store and process various data based on the received meteorological data and hydrological data, and report various forecast results to the early warning reporting module.

[0172] Optionally, the early warning reporting module includes:

[0173] A forecast data reporting module is used to report the predicted precipitation forecast data sequence and flow forecast data sequence;

[0174] The disaster warning module is used to issue a warning when the predicted flow forecast data sequence is analyzed to determine the existence of disaster risks.

[0175] In this embodiment, the forecast data reporting module collects the precipitation forecast data sequence obtained by the precipitation forecast module and the flow forecast data sequence obtained by the hydrological forecast module, and reports the forecast results to the unified dispatching center.

[0176] Optionally, the system 300 further includes an intelligent light storage power module, which includes a solar charging control module, a short message communication module, a battery and a controller module;

[0177] The solar charging control module is used to convert the voltage generated by the solar panel into a battery charging voltage;

[0178] SMS communication module, used to receive remote commands sent by the user end at regular intervals and respond;

[0179] Batteries are used to store and release electrical energy to power the meteorological and hydrological forecasting systems;

[0180] The controller module is used to detect the power storage state of the battery according to the instruction information of the user end, and to control the power transmission between the solar charging control module and the battery, and to control the sleep and wake-up of the intelligent power supply.

[0181] Optionally, the precipitation forecast module 301 includes:

[0182] A simulation unit, used to simulate climate data through various parameterization schemes of the weather forecast model and determine the initial precipitation forecast data sequence of the target basin under various parameterization schemes;

[0183] A representative scheme determination unit is used to calculate the initial precipitation forecast data sequence under various parameterized schemes through a parameterized scheme screening algorithm to determine the target number of representative schemes;

[0184] The fusion unit is used to fuse the initial precipitation forecast data sequence of the current date of the target basin under each representative scheme through a fusion algorithm to obtain the precipitation forecast data sequence of the target basin.

[0185] Optionally, the representative solution determining unit includes:

[0186] The average precipitation difference data sequence determination unit is used to determine the average precipitation difference data sequence according to the initial precipitation forecast data sequence and the measured precipitation data sequence of all parameterized schemes;

[0187] A first determining unit is used to determine a parameterization scheme corresponding to a precipitation difference data sequence closest to the average precipitation difference data sequence as a first representative scheme;

[0188] A second determining unit is used to determine the parameterized scheme corresponding to the precipitation difference data sequence farthest from the precipitation difference data sequence of the first representative scheme as the second representative scheme;

[0189] A distance calculation unit, used to calculate the first distance between the precipitation difference data series of the remaining single parameterized schemes and the precipitation difference data series of each representative scheme;

[0190] a distance determining unit, configured to determine the smallest first distance among all first distances corresponding to the remaining single parameterized schemes as the target distance of the single parameterized scheme;

[0191] The representative scheme determination subunit is used to determine the parameterized scheme corresponding to the largest target distance as the next representative scheme according to the target distances of the remaining parameterized schemes, so as to obtain representative schemes of the target number.

[0192] Optionally, the fusion unit is used to obtain the precipitation forecast data sequence of the target basin by bringing the initial precipitation forecast data sequence of the current date of the target basin under each representative scheme into the fusion algorithm for calculation. The expression of the fusion algorithm is:

[0193]

[0194] Among them, p[y(t)|O(t)] represents the precipitation forecast data of the tth day in the precipitation forecast data sequence, p[y(t)|f i (t), O(t)] represents the initial precipitation forecast data of the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, w i (t) represents the weight value corresponding to the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, and n represents the total number of representative schemes.

[0195] Optionally, the hydrological forecast module 302 includes:

[0196] A first input unit is used to input the meteorological and hydrological data of the target meteorological and hydrological data type within the pre-hysteresis period into each target hydrological forecast sub-model in the target hydrological forecast model according to the pre-hysteresis period of the target meteorological and hydrological data type of the target basin;

[0197] A second input unit is used to input the precipitation forecast data into a target hydrological forecast sub-model corresponding to the forecast time according to the forecast time corresponding to the precipitation forecast data in the precipitation forecast data sequence;

[0198] A first hydrological forecasting unit, configured to predict, based on input data, the flow forecast data at the respective corresponding forecast times by the respective target hydrological forecasting sub-models;

[0199] The second hydrological forecasting unit is used to determine all flow forecast data obtained by predicting all target hydrological forecasting sub-models as a flow forecast data sequence for the target river basin.

[0200] Optionally, the system further comprises a target type determination unit, configured to determine a target meteorological and hydrological data type;

[0201] The target type determination unit includes:

[0202] A first data sequence determining unit, used to determine a first monthly data sequence of the flow of the target basin within a first time period starting from a first moment;

[0203] A second data sequence determination unit, used to determine a second monthly data sequence of meteorological and hydrological data types within the first time period starting from a second moment, wherein the second moment is a moment that is separated from the first moment by a preset time period and is before the first moment;

[0204] A correlation calculation unit, configured to obtain a correlation value corresponding to the preset time interval by performing correlation calculation on the first month data sequence and the second month data sequence;

[0205] A first comparison unit is used to compare the correlation values ​​between the flow data and the meteorological and hydrological data types at different preset time intervals with set thresholds to obtain a comparison result;

[0206] The target type determination subunit is used to determine the meteorological and hydrological data type as the target meteorological and hydrological data type when there is an absolute value of a correlation value greater than or equal to a set threshold among all the absolute values ​​of the correlation values ​​represented by the comparison result.

[0207] Optionally, the system further comprises a pre-hysteresis determination unit, configured to determine a pre-hysteresis of a target meteorological and hydrological data type;

[0208] The pre-delay period determination unit comprises:

[0209] A correlation value determination unit, used to determine the maximum correlation value absolute value among the correlation value absolute values ​​of the flow rate and the target meteorological and hydrological data type at different preset time intervals;

[0210] The pre-hysteresis determination subunit is used to determine the preset time interval corresponding to the absolute value of the maximum correlation value as the pre-hysteresis of the target meteorological and hydrological data type.

[0211] Optionally, the system further includes:

[0212] A second comparison unit is used to compare the flow forecast data sequence with the flood extreme value to obtain a first comparison result;

[0213] A first early warning unit, configured to determine a target forecast time for a flood event to occur according to the first comparison result, and to issue a corresponding early warning;

[0214] A third comparison unit is used to compare the flow forecast data sequence with the low water extreme value to obtain a second comparison result;

[0215] The second early warning unit is used to determine the target forecast time of the drought event according to the second comparison result, and to issue a corresponding early warning.

[0216] As for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0217] It should be noted that, for the method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present application are not limited by the described order of actions, because according to the embodiments of the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present application.

[0218] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0219] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the embodiments of the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0220] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0221] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0222] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0223] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.

[0224] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0225] The above is a detailed introduction to a method and system for meteorological and hydrological forecasting provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for meteorological and hydrological forecasting, characterized in that: The method comprises: Determine the precipitation forecast data sequence of the target basin based on the climate data of the target basin; Inputting the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin; The precipitation forecast data sequence and the flow forecast data sequence are superimposed on the constructed three-dimensional digital model, so as to perform three-dimensional visualization of the forecast data sequence of the target watershed through the three-dimensional digital model.

2. A method for meteorological and hydrological forecasting according to claim 1, characterized in that: According to the climate data of the target basin, the precipitation forecast data series of the target basin is determined, including: The climate data are simulated through various parameterization schemes of the weather forecast model to determine the initial precipitation forecast data series of the target basin under various parameterization schemes; The initial precipitation forecast data series under various parameterized schemes are calculated through the parameterized scheme screening algorithm to determine the representative schemes of the target number; The initial precipitation forecast data series of the target basin on the current date under each representative scheme are fused through the fusion algorithm to obtain the precipitation forecast data series of the target basin.

3. A method for meteorological and hydrological forecasting according to claim 2, characterized in that: The initial precipitation forecast data series under various parameterized schemes are calculated through the parameterized scheme screening algorithm to determine the representative schemes of the target number, including: According to the initial precipitation forecast data series and the measured precipitation data series of all parameterization schemes, the average precipitation difference data series is determined; Determine the parameterization scheme corresponding to the precipitation difference data sequence closest to the average precipitation difference data sequence as the first representative scheme; Determine the parameterized scheme corresponding to the precipitation difference data sequence farthest from the precipitation difference data sequence of the first representative scheme as the second representative scheme; Calculate the first distance between the precipitation difference data series of the remaining single parameterization schemes and the precipitation difference data series of each representative scheme; Determine the smallest first distance among all first distances corresponding to the remaining single parameterized solutions as the target distance of the single parameterized solution; According to the target distances of the remaining parameterized solutions, the parameterized solution corresponding to the largest target distance is determined as the next representative solution to obtain a representative solution for the target number.

4. A method for meteorological and hydrological forecasting according to claim 2, characterized in that: The initial precipitation forecast data sequence of the target basin on the current date under each representative scheme is fused through the fusion algorithm to obtain the precipitation forecast data sequence of the target basin, including: By bringing the initial precipitation forecast data sequence of the target basin on the current date under each representative scheme into the fusion algorithm for calculation, the precipitation forecast data sequence of the target basin is obtained. The expression of the fusion algorithm is: Among them, p[y(t)|O(t)] represents the precipitation forecast data of the tth day in the precipitation forecast data sequence, p[y(t)|f i (t), O(t)] represents the initial precipitation forecast data of the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, w i (t) represents the weight value corresponding to the tth day in the initial precipitation forecast data sequence under the i-th representative scheme, and n represents the total number of representative schemes.

5. The method for meteorological and hydrological forecasting according to claim 1, characterized in that: The precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period are input into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin, including: According to the pre-hysteresis period of the target meteorological and hydrological data type of the target basin, the meteorological and hydrological data of the target meteorological and hydrological data type within the pre-hysteresis period are input into each target hydrological forecast sub-model in the target hydrological forecast model; According to the forecast time corresponding to the precipitation forecast data in the precipitation forecast data sequence, inputting the precipitation forecast data into the target hydrological forecast sub-model corresponding to the forecast time; According to the input data, each target hydrological forecast sub-model predicts and obtains the flow forecast data at the corresponding forecast time; All flow forecast data obtained by predicting all target hydrological forecast sub-models are determined as the flow forecast data sequence of the target basin.

6. A method for meteorological and hydrological forecasting according to claim 5, characterized in that: Determine the target meteorological and hydrological data types, including: Determine the first month data sequence of the flow of the target basin within the first time period starting from the first moment; Determine a second month data sequence of meteorological and hydrological data types starting from a second moment within the first time period, where the second moment is a moment that is separated from the first moment by a preset time period and is before the first moment; By performing correlation calculation on the first month's data sequence and the second month's data sequence, a correlation value corresponding to the preset time interval is obtained; The comparison result is obtained by comparing the absolute values ​​of the correlation values ​​of the meteorological and hydrological data types at different preset time intervals with the set thresholds respectively; When the comparison result indicates that there is an absolute value of a correlation value greater than or equal to a set threshold value among all absolute values ​​of the correlation values, the meteorological and hydrological data type is determined as the target meteorological and hydrological data type.

7. A method for meteorological and hydrological forecasting according to claim 6, characterized in that: Determine the lead time for the target meteorological and hydrological data types, including: Determine the maximum absolute value of the correlation value between the flow and the target meteorological and hydrological data type at different preset time intervals; The preset time interval corresponding to the absolute value of the maximum correlation value is determined as the front lag period of the target meteorological and hydrological data type.

8. A method for meteorological and hydrological forecasting according to claim 6, characterized in that: The method further comprises: Comparing the flow forecast data series with the flood extreme value to obtain a first comparison result; Determine the target forecast time of a flood event according to the first comparison result, and issue a corresponding warning; Compare the flow forecast data series with the low water extreme value to obtain a second comparison result; According to the second comparison result, the target forecast time of the drought event is determined, and a corresponding warning is issued.

9. A method for meteorological and hydrological forecasting according to claim 6, characterized in that: Meteorological and hydrological data types include: precipitation, evaporation, flow, and soil moisture.

10. A method for meteorological and hydrological forecasting according to claim 5, characterized in that: The target hydrological forecast model is a bidirectional long short-term memory neural network.

11. A method for meteorological and hydrological forecasting according to claim 9, characterized in that: According to the first comparison result, the target forecast time of the flood event is determined, and a corresponding early warning is issued, including: When the first comparison result is that there is flow forecast data greater than or equal to the flood extreme value in the flow forecast data sequence, the forecast time corresponding to each flow forecast data greater than or equal to the flood extreme value is determined as the target forecast time, and a corresponding warning is issued; Determining the target forecast time of the drought event according to the second comparison result and issuing a corresponding warning includes: When the second comparison result is that the flow forecast data of one day in the flow forecast data sequence and the daily flow average of multiple days before that day are less than the low water extreme value, that day is determined as the target forecast time and a corresponding warning is issued.

12. A system for meteorological and hydrological forecasting, characterized in that: The system includes a precipitation forecast module, a hydrological forecast module and a body digital module; The precipitation forecast module is used to determine the precipitation forecast data sequence of the target basin according to the climate data of the target basin; The hydrological forecast module is used to input the precipitation forecast data sequence and the meteorological and hydrological data of the target basin in the pre-lag period into the target hydrological forecast model for processing to obtain the flow forecast data sequence of the target basin; The main body digital module is used to superimpose the precipitation forecast data sequence and the flow forecast data sequence with the constructed three-dimensional digital model, so as to perform three-dimensional visualization of the forecast data sequence of the target watershed through the three-dimensional digital model.

13. A meteorological and hydrological forecasting system according to claim 12, characterized in that: The main body digital module includes: A watershed data module, used for constructing the three-dimensional digital model based on the geographic data of the target watershed; The meteorological and hydrological monitoring module is used to monitor and obtain the meteorological and hydrological data of the target basin.

14. A meteorological and hydrological forecasting system according to claim 13, characterized in that: The system also includes a data interaction module and an early warning reporting module; The data interaction module is used to store meteorological and hydrological data, and send the predicted precipitation forecast data sequence and flow forecast data sequence to the early warning reporting module for alarm.

15. A meteorological and hydrological forecasting system according to claim 14, characterized in that: The early warning reporting module includes: A forecast data reporting module is used to report the predicted precipitation forecast data sequence and flow forecast data sequence; The disaster warning module is used to issue a warning when the predicted flow forecast data sequence is analyzed to determine the existence of disaster risks.

16. A meteorological and hydrological forecasting system according to claim 11, characterized in that: The system also includes an intelligent light storage power module, which includes a solar charging control module, a short message communication module, a battery and a controller module; The solar charging control module is used to convert the voltage generated by the solar panel into a battery charging voltage; SMS communication module, used to receive remote commands sent by the user end at regular intervals and respond; Batteries, used for storing and releasing electrical energy; The controller module is used to detect the power storage state of the battery according to the instruction information of the user end, and to control the power transmission between the solar charging control module and the battery, and to control the sleep and wake-up of the intelligent power supply.

Images