A regional extreme precipitation forecast and warning method based on cGAN

Abstract

A regional extreme precipitation forecasting and warning method based on a conditional generative adversarial network (cGAN) involves preprocessing multi-source meteorological data, including standardization and spatial consistency alignment, to ensure data usability. The cGAN encoding module extracts and fuses interannual daily mean and anomaly features to construct a multi-scale feature pyramid, enhancing the representation of rainfall spatial modalities. Forecast data is generated using the cGAN generator and conditional discriminator. The discriminator incorporates a rainfall synergy factor as a physical constraint, and network parameters are optimized using an adversarial loss function and a physical consistency loss. Forecast data is post-processed using a graph neural network (GNN), using meteorological station observations as the ground truth for spatial dependency modeling and optimization correction, significantly improving the forecast accuracy and reliability of extreme precipitation events.

Description

Technical Field

[0001] The present invention relates to precipitation modeling and prediction technology, and in particular to a regional extreme precipitation forecasting and early warning method based on cGAN. Background Art

[0002] Precipitation modeling and prediction methods are diverse, encompassing conceptual models, physical models, numerical methods, and statistical approaches. In recent years, artificial intelligence technologies, particularly deep generative models (DGMs), have been widely studied in the field of rainfall forecasting. Generative adversarial networks (GANs), a typical deep generative model, transform complex likelihood functions into a neural network framework, enabling the model to fit the target distribution by optimizing parameters. The core concept of GANs is to use adversarial training between a generator and a discriminator to enable the generator to produce samples that are closer to real data.

[0003] In the field of precipitation forecasting, Jing et al. designed a GAN-based precipitation forecast model in which the generator uses a ConvLSTM structure and is equipped with two discriminators for radar extrapolation. The generator loss function of this model combines the mean squared error (MSE) and mean absolute error (MAE), while the discriminator uses the binary cross entropy (BCE) loss. The research team constrained the radar reflectivity between 0 and 75 decibels and successfully predicted high-resolution radar data for 1.5 hours in the future. Furthermore, Ravuri et al. proposed another GAN-based precipitation forecast model, DGMR, which uses a ConvGRU as the generator and two discriminators to capture spatial and temporal patterns, respectively. Relying solely on radar observations, the DGMR model successfully achieved high-resolution precipitation forecasts and outperformed other models in a critical success index (CSI) assessment. Specifically targeting heavy precipitation events, this algorithm utilizes a hinge loss, with the discriminator using the hinge loss and the generator combining the hinge loss with the MAE.

[0004] GANs have achieved remarkable success in super-resolution and statistical downscaling, particularly in short-term forecasting, effectively alleviating the problem of smoothing and loss of intensity over time in forecasts. For example, Ravuri et al. (2021) used GANs to predict global precipitation probabilistically, improving the spatial resolution of the forecasts while preserving the structural characteristics of extreme precipitation events. Price and Rasp (2022) used conditional GANs (cGANs) to downscale numerical weather predictions (NWPs), significantly improving the clarity of small-scale precipitation structures and reducing forecast uncertainty.

[0005] In summary, the application of GANs in precipitation modeling and prediction has demonstrated its unique advantages in capturing complex precipitation patterns and extreme events, providing a new technical approach for high-resolution precipitation forecasting and accurate prediction of extreme weather events.

[0006] Although GANs have shown great potential for precipitation forecast correction and downscaling, current GAN-based precipitation forecasting research rarely considers precipitation data from meteorological stations. However, precipitation data from meteorological stations generally more accurately reflects actual surface precipitation and is crucial for operational precipitation forecasting, particularly for responding to emergencies caused by heavy rainfall. Furthermore, existing models have a low reliance on multi-source data and physical constraints, which may limit their ability to effectively capture the incremental information generated by the fusion of different information sources and their physical consistency with actual rainfall, thus affecting forecast accuracy.

[0007] It should be noted that the information disclosed in the above background technology section is only used to understand the background of this application, and therefore may include information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0008] The main purpose of the present invention is to overcome the defects existing in the above-mentioned background technology and provide a regional extreme precipitation forecasting and warning method based on cGAN.

[0009] To achieve the above object, the present invention adopts the following technical solutions:

[0010] A regional extreme precipitation forecasting and warning method based on a conditional generative adversarial network (cGAN) includes the following steps:

[0011] S1. Preprocessing of multi-source meteorological data: standardizing multi-source meteorological data to align spatial consistency of data of different resolutions. The multi-source meteorological data includes rainfall and its synergistic factors.

[0012] S2. Feature extraction and fusion: The preprocessed data is input into the encoding module of the conditional generative adversarial network (cGAN). The interannual daily mean features and anomaly features are extracted through a multi-branch structure. Multi-source features are fused based on the attention mechanism, and a multi-scale feature pyramid is constructed to enhance the representation of rainfall spatial modalities.

[0013] S3. Generate forecast data: The fused features are input into the generator of a conditional generative adversarial network (cGAN), and forecast data is generated through a decoding module. The generated data is supervised by a conditional discriminator. The discriminator explicitly introduces the rainfall synergy factor as a physical constraint and optimizes the network parameters using an adversarial loss function and a physical consistency loss to ensure that the generated data conforms to meteorological physical laws.

[0014] S4. Graph neural network post-processing: A topological map is constructed based on the coordinates of meteorological stations and grid points. The meteorological station observations are used as the true values. The generated forecast data is spatially dependently modeled and optimized using a graph neural network (GNN) to improve the forecast accuracy of extreme precipitation events.

[0015] A computer-readable storage medium stores a computer program, which implements the method when executed by a processor.

[0016] A computer program product comprises a computer program, which implements the method when executed by a processor.

[0017] The present invention has the following beneficial effects:

[0018] This paper proposes a regional extreme precipitation forecasting and early warning method based on a conditional generative adversarial network (CGAN). This method integrates multi-source data and strengthens the model's ability to model physical laws, significantly improving the forecast accuracy and reliability of extreme precipitation events. The integration of multi-source data provides the model with greater information gain, enabling it to more effectively capture incremental information between different sources and enhancing the consistency between forecast data and actual precipitation. By using meteorological synergy factors as conditional information, the CGAN model is forced to learn the underlying nonlinear relationship between rainfall and synergy factors. This not only improves forecast accuracy but also enhances the physical consistency and reliability of the forecast results. Furthermore, by leveraging the powerful modeling capabilities of graph neural networks (GNNs) for complex spatial dependencies, combined with the topological structure between meteorological stations and grid points, the method utilizes a neighboring information propagation mechanism to dynamically update node features, effectively capturing local details and global spatial variations. This method performs post-processing corrections on the initial CGAN forecast results, significantly reducing systematic errors, enhancing the model's ability to represent extreme precipitation events, effectively mitigating oversmoothing in heavy precipitation forecasts, and improving the precision and reliability of forecast results. Compared with traditional GAN models, this invention performs better in heavy precipitation forecasting. Through a two-step forecasting method, preliminary forecast results are first generated through cGAN, and then the spatial information transmission between GNN learning nodes is used to correct the preliminary forecast results in combination with high-precision meteorological observation data, thereby significantly improving the forecast accuracy and reliability of extreme precipitation events.

[0019] Compared with the prior art, the main advantages of the present invention are:

[0020] 1. Improving Data Diversity and Physical Modeling: Traditional GAN models do not explicitly consider the differences in learning between long-term and short-term rainfall patterns, lacking physical meaning. They only capture structural and intensity information by optimizing the loss function, failing to fully consider the impact of synergistic factors on rainfall forecasts. This paper innovatively designs a dual-branch learning structure for long-term cyclical patterns and short-term fluctuations, capturing long-term trends and short-term variations respectively, which is more consistent with meteorological physics. Furthermore, using a conditional GAN (cGAN) framework, rainfall synergistic factors (such as temperature, humidity, and wind speed) are input as conditional inputs to the generator while also considering the constraints of the synergistic factors in the discriminator. This forces the model to focus on and learn the underlying nonlinear relationship between rainfall and synergistic factors, thereby generating more accurate precipitation forecasts that are more consistent with physical laws.

[0021] 2. Improved Heavy Precipitation Forecast Accuracy: Traditional GAN models often exhibit oversmoothing in heavy precipitation forecasts, resulting in insufficient prediction capabilities for extreme precipitation events. This paper proposes a two-step forecasting approach: first, a cGAN is used to generate preliminary forecasts. Then, a graph neural network (GNN) is used to learn the spatial information transfer between nodes, and the preliminary forecasts are corrected using high-precision meteorological observation data. This approach effectively mitigates the oversmoothing effect of cGAN on heavy precipitation, significantly improving the accuracy and reliability of extreme precipitation forecasts.

[0022] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 Flowchart of a regional extreme precipitation forecast and warning method according to an embodiment of the present invention.

[0024] Figure 2 The network architecture of cGAN and GNN in the embodiments of the present invention.

[0025] Figure 3 A multi-scale feature extraction module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope of the present invention and its application.

[0027] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0028] This paper addresses technical issues in existing rainfall forecast models, such as insufficient ability to identify spatiotemporal features, prone to oversmoothing in extreme precipitation forecasts, and limited forecast accuracy. By doing so, it proposes an intelligent extreme precipitation forecasting and warning method and system based on a Conditional Generative Adversarial Network (cGAN). This invention significantly improves the spatiotemporal forecast accuracy and warning reliability of extreme precipitation events through an innovative two-step forecasting framework.

[0029] The precipitation forecast and correction method based on cGANs focuses on multi-source data fusion and stronger spatiotemporal feature extraction, as well as strengthening the physical constraints of the model to improve the accuracy of heavy rainfall forecasts. Specifically, it can combine multi-source meteorological data such as satellite, radar, and reanalysis data, and capture the complex spatiotemporal evolution characteristics of the precipitation system through the generator and discriminator architecture of cGANs. At the same time, meteorological observation station data is introduced to further correct the model prediction results to enhance the consistency between the forecast data and actual precipitation, and reduce the over-smoothing effect of the model on extreme precipitation events. Through this method, it aims to improve the reliability of precipitation forecasts, provide strong support for accurate meteorological services, and play an important role in the early warning and disaster prevention and mitigation of heavy rainfall events.

[0030] See Figures 1 to 3 The embodiment of the present invention provides a regional extreme precipitation forecasting and warning method based on a conditional generative adversarial network (cGAN), comprising the following steps:

[0031] S1. Preprocessing of multi-source meteorological data: standardizing multi-source meteorological data to align spatial consistency of data of different resolutions. The multi-source meteorological data includes rainfall and its synergistic factors.

[0032] In some embodiments, step S1 specifically includes: using an adaptive normalization method to uniformly process multi-source meteorological data (e.g., satellite, radar, and reanalysis data), including the spatiotemporal alignment of historical precipitation data; using a differentiable interpolation algorithm to spatially consistent the standardized multi-source meteorological data, interpolating data of varying resolutions to a uniform spatial resolution; and adding random noise perturbations to the standardized multi-source meteorological data to enhance the model's robustness to data uncertainty and noise. Furthermore, in step S1, rainfall data and its synergistic factors at multiple time steps (t1, t2, ..., tn) are used as model inputs. The synergistic factors include variables such as wind field, temperature, humidity, and air pressure. ERA5 data is used as the true value source to construct a time series dataset for model training and prediction.

[0033] S2. Feature extraction and fusion: The preprocessed data is input into the encoding module of the conditional generative adversarial network (cGAN). The interannual daily average features and anomaly features are extracted through a multi-branch structure. Multi-source features are fused based on the attention mechanism, and a multi-scale feature pyramid is constructed to enhance the representation ability of rainfall spatial modalities.

[0034] In some embodiments, step S2 specifically includes: inputting the preprocessed multi-source meteorological data into the encoder of the cGAN, mapping the data to a spatial latent representation, and regularizing the latent space using the KL divergence; uniformly processing the rainfall and its synergistic factor data of different spatial resolutions through a differentiable interpolation layer, which is composed of a convolutional layer and a pooling layer to ensure the spatial consistency of the data; decomposing the data into interannual average daily data and anomaly data, and using the spatiotemporal feature extraction module of the Conv-Transformer combination to extract features of the two types of data respectively; splicing the interannual average daily features and anomaly features of multiple data through the concat layer, calculating the attention of each feature using the attention mechanism, and fusing the features based on the attention score; performing spatial multi-scale feature extraction on the fused features to enhance the representation ability of semantic information and fine-grained information.

[0035] S3. Generate forecast data: The fused features are input into the generator of the conditional generative adversarial network (cGAN), and forecast data is generated through the decoding module. The generated data is supervised by a conditional discriminator. The discriminator explicitly introduces the rainfall synergy factor as a physical constraint condition and optimizes the network parameters through the adversarial loss function and physical consistency loss to ensure that the generated data conforms to the laws of meteorological physics.

[0036] In some embodiments, step S3 specifically includes: inputting the enhanced multi-scale spatial features into the decoding module of the cGAN generator, mapping the potential features into the rainfall and its co-factor data at the next moment through layer-by-layer convolution operations; in the decoding module, using the attention mechanism and feature concatenation (concat) to further fuse the multi-scale spatial feature information to predict the rainfall and its co-factor data at the next moment; inputting the forecast data generated by the generator into the discriminator, and the discriminator determines whether the generated data comes from the real data distribution by calculating the discriminant loss Loss, further constraining and optimizing the output of the generator, and improving its modeling ability for real rainfall data; and inputting the generated forecast data as new input into the trained cGAN model to achieve multi-step prediction.

[0037] S4. Graph neural network post-processing: A topological map is constructed based on the coordinates of meteorological stations and grid points. The meteorological station observations are used as the true values. The generated forecast data is spatially dependently modeled and optimized using a graph neural network (GNN) to improve the forecast accuracy of extreme precipitation events.

[0038] In some embodiments, in step S4, a topological map is constructed based on the coordinates of the meteorological station and the ERA5 data grid points to characterize the spatial relationship; the ERA5 rainfall data and the data interpolated to the meteorological station are used as inputs of a graph neural network (GNN); the observation values of the meteorological station are used as the true values, and the GNN network is trained as a post-processing module to perform spatial dependency modeling and optimization correction on the generated forecast data to improve the forecast accuracy.

[0039] In some embodiments, in step S4, a forecast correction module is used to map the forecast data generated by cGAN to the location of the meteorological station through interpolation, and optimize the difference with the real station data. Combined with the GNN spatial information transmission mechanism, the rainfall data between adjacent grid points is adjusted to achieve spatial detail enhancement of the generated regular forecast data, so as to solve the over-smoothing problem of extreme rainfall forecasts and make the generated data more consistent with the actual surface observation rainfall data.

[0040] In some embodiments, step S4 specifically includes the following steps: interpolating the forecast data generated by cGAN to the spatial position of the meteorological station, and mapping the forecast data to each meteorological station using an interpolation function; merging the interpolated forecast data with the original forecast data as the input of the graph neural network (GNN); constructing a graph structure of the graph neural network, in which the nodes include meteorological stations and cGAN forecast data grid points, and the edges are constructed according to the distance between the stations; inputting the node feature matrix and the adjacency matrix into the GNN, capturing the spatial dependency and generating optimized forecast data through message passing, message aggregation, node update and multi-layer propagation mechanism; using the meteorological station observation value as the true value, calculating the mean square error (MSE) loss, and updating the GNN network parameters through back propagation to achieve optimized correction of the forecast data.

[0041] In the present invention, multi-source meteorological data preprocessing provides standardized input for feature extraction and fusion, which in turn provides enhanced features for generating forecast data. The generated forecast data generates preliminary results through cGAN, and graph neural network post-processing optimizes and corrects the preliminary results, ultimately achieving high-precision extreme precipitation forecast and warning.

[0042] The following further describes specific embodiments of the present invention and examples of algorithm implementation thereof.

[0043] This paper proposes a regional extreme precipitation forecast and warning method based on the conditional generative adversarial network (cGAN). It constructs a multi-source meteorological data feature fusion framework based on the attention mechanism, adds a rainfall synergy factor and a physical consistency discriminator of rainfall, and adds a graph neural network (GNN) optimization module, which significantly improves the forecast accuracy and reliability of extreme precipitation events.

[0044] In one embodiment of the present invention, an adaptive normalization method is first used to uniformly process multi-source meteorological data, including the spatiotemporal alignment of historical precipitation data, and a differentiable interpolation algorithm is used to achieve spatial consistency of data with different resolutions.

[0045] In terms of model architecture, an embodiment of the present invention proposes a feature fusion cGAN network based on an attention mechanism, which adopts a multi-branch structure to capture the interannual daily average characteristics and anomaly characteristics, and constructs a multi-scale feature pyramid to enhance the representation ability of rainfall spatial modalities. At the same time, a rainfall synergy factor and a physical consistency discriminator of rainfall are constructed to enhance physical constraints. In addition, a two-step mechanism is innovatively introduced to further improve the forecast accuracy. Specifically, in order to make the forecast results more consistent with the actual meteorological observation results and solve the over-smoothing problem of extreme rainfall events, the present invention constructs a topological map based on the coordinates of meteorological stations and grid points, takes the reanalysis data and its interpolation results at the observation station as input, takes the meteorological station observation value as the true value, and uses the graph neural network (GNN) to further optimize the forecast results, thereby improving the accuracy of the forecast.

[0046] The present invention adopts the following main steps (see Figure 1 ):

[0047] 1) Multi-source meteorological data preprocessing: Random noise perturbation is added to all input standardized multi-source meteorological data, and data of different resolutions are aligned using a differentiable interpolation algorithm to ensure spatial consistency and reduce spatial bias.

[0048] 2) Feature Extraction and Fusion: The preprocessed data is fed into the cGAN encoding module. Specifically, a multi-branch structure is used to extract interannual daily average and anomaly features. Multi-source features are then fused using an attention mechanism to enhance the representation of multi-source data. A multi-scale feature extraction module is constructed to enhance the spatial modal representation of rainfall.

[0049] 3) Generate forecast data: Input the encoded features into the decoder module of the cGAN network, and enhance the decoder's generation ability through the conditional discriminator.

[0050] 4) cGNN post-processing: A topological map is constructed based on the coordinates of the meteorological stations and ERA5 data grid points. The ERA5 rainfall data and the data interpolated to the meteorological stations based on ERA5 are used as input. The meteorological station observations are used as the true values, and a GNN network is trained as a post-processing module to optimize the cGAN forecast results and improve their accuracy and reliability.

[0051] In steps 1) and 2), the present invention first uses a differentiable interpolation algorithm to interpolate standardized, random noise-perturbed, multi-source meteorological data to a uniform spatial resolution to ensure spatial consistency. Subsequently, preliminary feature extraction is performed on the multi-source data, including interannual daily averages and anomalies. An attention mechanism is then used to fuse features from different data sources. Spatial scale feature extraction is then performed on these fused features to enhance feature representation capabilities.

[0052] In step 3), the present invention generates forecast data by decoding the encoded features. To verify the accuracy of the generated forecast data, the present invention calculates a loss between the forecast data and the true value, and uses the cGAN discriminator to calculate the discriminant loss, and then reversely optimizes the network parameters. Simultaneously, the conditional discriminator supervises the generator output, ensuring the accuracy of the generated data while also ensuring that it conforms to physical laws.

[0053] In step 4), in order to make the generated data more consistent with the actual surface rainfall data and to solve the over-smoothing problem of the forecast model for extreme rainfall conditions, the present invention adds a forecast correction module after the cGAN forecast model to provide a more detailed spatial description of the actual rainfall data.

[0054] The specific method flow of the present invention is as follows Figure 1 As shown, the network structures of cGAN and GNN are as follows Figure 2 shown.

[0055] The present invention can be divided into two stages.

[0056] The first stage is forecast data generation, and the specific steps are as follows:

[0057] In the forward forecasting work based on cGAN, the present invention takes the rainfall and its synergistic factor data at time t1, t2, ..., tn as the input of the model, and takes the rainfall and its synergistic factor (such as wind field, temperature, humidity, air pressure and other variables, using ERA5 data) at time t as the true value, that is, .

[0058] cGAN consists of a generator G and a discriminator D. The input data first passes through the trained encoder ɛ of G, which maps the multi-source data to the spatial potential representation Z, that is, , KL divergence will be used to regularize the latent space. The multi-source data input layer will first pass through a differentiable interpolation layer, which is mainly composed of a convolution (Conv) and a pooling layer (Pool), to unify the rainfall and its synergistic factor data of different spatial resolutions to the same spatial scale. Secondly, each layer of data will be decomposed into inter-annual average daily data and anomaly data. The corresponding features are preliminarily extracted through the spatiotemporal feature extraction module of the Conv-transformer combination. The two features of multiple data are preliminarily spliced through the concat layer, and the attention mechanism method is used to calculate the attention of each feature, and these features are fused based on the attention score. Then, the spatial multi-scale feature extraction operation is performed on the preliminarily fused features (see Figure 3 ), in order to obtain features that have stronger representation capabilities for semantic information and fine-grained information.

[0059] Then, the enhanced multi-scale spatial features are input into the decoding module γ of the subsequent cGAN G. Through layer-by-layer convolution operations, the potential features are mapped to the rainfall and its synergistic factor data at the next moment, that is, Specifically, in the decoding module, the extracted multi-scale spatial feature information is further spliced and fused through the attention mechanism and concat, and the data of rainfall and its synergistic factors at time t+1 are predicted through layer-by-layer convolution operations. The discriminator D is to judge the (X t+1 ,Y t+1 )_predict, whether it comes from the real data distribution (X t+1 ,Y t+1 )_true, so the decoder generates (X t+1 ,Y t+1 )_predict is input into D to calculate the discriminant loss (Loss_D = D((X t+1 ,Y t+1 )_predict,(X t+1 ,Y t+1 )_true)), further constraining and optimizing the generator's output to improve its ability to model real rainfall data. Furthermore, the predicted data at time t+1 is fed into the trained cGAN model to generate the forecast data for time t+2, enabling multi-step prediction.

[0060] Among them, the noise z is a random vector sampled from a prior distribution (such as a standard normal distribution); the conditional information y is the input data y= (x r t , x c t );G r(z|y) is the rainfall data generated at time t+1; G c (z|y) is the generated meteorological synergy factor at time t+1.

[0061] The second stage is forecast post-processing. The specific steps are as follows:

[0062] Specifically, the trained cGAN is used to generate the rainfall forecast data G at time t+1 r (z|y). The generated forecast data G r (z|y) is interpolated to the spatial location of the meteorological station. Assume that the location coordinates of the meteorological station are (S i ) i=1:N , where S i is the coordinate of the ith station, using the Kriging interpolation function Interpolate forecast data to station locations: The interpolated forecast data is combined with the original forecast data as input data.

[0063] The merged data is input into the trained graph neural network (GNN) to generate post-processed forecast data. The graph structure of the GNN network is as follows: 1) Node: Each meteorological station S i The node corresponding to the cGAN prediction data grid point has the node feature X i t+1 ;2) Edge: construct the edge according to the distance between sites, and the adjacency matrix is A. The input of GNN is the node feature matrix X={X i} i=1:N And the adjacency matrix A. The output of GNN is the post-processed forecast data. The present invention uses the meteorological station observation value as the true value, calculates the MSE loss, and gradually updates the GNN network parameters: In GNN, proximity information propagation is one of the core mechanisms, which transmits and aggregates information between nodes through the graph structure (nodes and edges). Specifically: (1) Message passing: For each node v i , from its neighbor nodes Collect information. The message function M defines how to get information from neighbor nodes v j Pass information to node v i (2)Message aggregation: For node v i Aggregate all neighbor information. (3) Node update: Use the update function U to combine the aggregated message mi with the node’s current feature Combined, the feature representation of the node in the L+1 layer is generated. (4) Multi-layer propagation: Through multi-layer information transmission, the node features gradually integrate the global information. (4) Output: The node features of the last layer It can be used as post-processed forecast data. Through multi-layer propagation, GNN can capture complex spatial dependencies and generate more reliable forecast results.

[0064] In summary, this paper innovatively constructs a multi-source meteorological data feature fusion framework based on the attention mechanism, adds a rainfall synergy factor and a rainfall physical consistency discriminator, and adds a graph neural network (GNN) optimization module, which significantly improves the forecast accuracy and reliability of extreme precipitation events.

[0065] The important features and innovative contributions of the present invention are:

[0066] 1. Multi-source data enhancement and robustness improvement: By introducing Gaussian noise into multi-source input data, the model's robustness to data uncertainty and noise is enhanced, improving the stability of extreme rainfall forecasts. At the same time, the fusion of multi-source data will bring more information gain.

[0067] 2. Conditional GAN modeling of nonlinear relationships: By using the conditional GAN model and incorporating meteorological synergistic factors (such as temperature, humidity, and wind speed) as conditional information, the model is forced to learn the potential nonlinear relationship between rainfall and synergistic factors, thereby improving forecast accuracy and enhancing the physical consistency and reliability of the forecast results.

[0068] 3. Observational Data-Driven Post-Processing Correction: Leveraging the powerful modeling capabilities of graph neural networks (GNNs) for complex spatial dependencies, this method incorporates the topological structure between meteorological stations and grid points and leverages neighboring information propagation to dynamically update node features, effectively capturing both local details and global spatial variations. This post-processing correction of the initial cGAN forecast results significantly reduces systematic errors, enhances the model's ability to represent extreme precipitation events, effectively mitigates oversmoothing in heavy precipitation forecasts, and improves the precision and reliability of forecasts.

[0069] Compared with the prior art, the present invention has the following significant advantages:

[0070] 1. Improving Data Diversity and Physical Modeling: Traditional GAN models do not explicitly consider the differences in learning between long-term and short-term rainfall patterns, lacking physical meaning. They only capture structural and intensity information by optimizing the loss function, failing to fully consider the impact of synergistic factors on rainfall forecasts. This paper innovatively designs a dual-branch learning structure for long-term cyclical patterns and short-term fluctuations, capturing long-term trends and short-term variations respectively, which is more consistent with meteorological physics. Furthermore, using a conditional GAN (cGAN) framework, rainfall synergistic factors (such as temperature, humidity, and wind speed) are input as conditional inputs to the generator while also considering the constraints of the synergistic factors in the discriminator. This forces the model to focus on and learn the underlying nonlinear relationship between rainfall and synergistic factors, thereby generating more accurate precipitation forecasts that are more consistent with physical laws.

[0071] 2. Improved Heavy Precipitation Forecast Accuracy: Traditional GAN models often exhibit oversmoothing in heavy precipitation forecasts, resulting in insufficient prediction capabilities for extreme precipitation events. This paper proposes a two-step forecasting approach: first, a cGAN is used to generate preliminary forecasts. Then, a graph neural network (GNN) is used to learn the spatial information transfer between nodes, and the preliminary forecasts are corrected using high-precision meteorological observation data. This approach effectively mitigates the oversmoothing effect of cGAN on heavy precipitation, significantly improving the accuracy and reliability of extreme precipitation forecasts.

[0072] An embodiment of the present invention further provides a storage medium for storing a computer program, which at least performs the above method when executed.

[0073] An embodiment of the present invention further provides a control device, comprising a processor and a storage medium for storing a computer program; wherein the processor is configured to execute at least the method described above when executing the computer program.

[0074] An embodiment of the present invention further provides a processor, which executes a computer program and at least performs the method described above.

[0075] The storage medium can be implemented by any type of non-volatile storage device, or a combination thereof. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); the magnetic surface memory can be a magnetic disk memory or a magnetic tape memory. The storage medium described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0076] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0077] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0078] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0079] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware associated with program instructions, and the aforementioned program may be stored in a computer-readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0080] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0081] The methods disclosed in the several method embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0082] The features disclosed in several product embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0083] The features disclosed in several method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0084] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. Those skilled in the art will recognize that, without departing from the scope of the present invention, several equivalent substitutions or obvious variations can be made, and the performance or use of the same should be considered to fall within the scope of protection of the present invention.

Claims

1. A regional extreme precipitation forecasting and warning method based on cGAN, characterized by: The following steps are involved: S1. Preprocessing of multi-source meteorological data: standardizing multi-source meteorological data to align spatial consistency of data of different resolutions. The multi-source meteorological data includes rainfall and its synergistic factors. S2. Feature extraction and fusion: The preprocessed multi-source meteorological data is input into the encoding module of the generative adversarial network (cGAN). The multi-branch structure is used to extract the interannual daily average features and anomaly features. The multi-source features are fused based on the attention mechanism, and a multi-scale feature pyramid is constructed to enhance the representation of rainfall spatial modalities. S3. Generate forecast data: The fused features are input into the generator of the conditional generative adversarial network (cGAN), and forecast data is generated through the decoding module. The generated data is supervised by a conditional discriminator. The discriminator explicitly introduces the rainfall synergy factor as a physical constraint and optimizes the network parameters through the adversarial loss function and physical consistency loss to ensure that the generated data conforms to the laws of meteorological physics. S4. Graph Neural Network Post-Processing: A topological map is constructed based on the coordinates of meteorological stations and grid points. The meteorological station observations are used as the true values. The generated forecast data is then modeled and optimized using a graph neural network (GNN), thereby improving the forecast accuracy of extreme precipitation events. Step S4 specifically includes the following steps: Interpolate the forecast data generated by cGAN to the spatial location of the meteorological station and use the interpolation function to map the forecast data to each meteorological station; The interpolated forecast data is combined with the original forecast data as the input of the graph neural network GNN; Construct a graph structure for the graph neural network, where nodes include meteorological stations and cGAN forecast data grid points, and edges are constructed based on the distance between stations; The node feature matrix and adjacency matrix are input into GNN, and spatial dependencies are captured and optimized forecast data is generated through message passing, message aggregation, node update and multi-layer propagation mechanisms; The meteorological station observations are taken as the true values, the mean square error (MSE) loss is calculated, and the GNN network parameters are updated through back propagation to achieve optimal correction of the forecast data.

2. The regional extreme precipitation forecasting and warning method based on cGAN according to claim 1 is characterized in that: Step S1 specifically includes: Adaptive normalization methods are used to uniformly process multi-source meteorological data, including the spatiotemporal alignment of historical precipitation data; The standardized multi-source meteorological data are processed for spatial consistency using a differentiable interpolation algorithm, and data with different resolutions are interpolated to a unified spatial resolution. Random noise perturbations are added to the standardized multi-source meteorological data to enhance the robustness of the model to data uncertainty and noise.

3. The regional extreme precipitation forecasting and warning method based on cGAN according to claim 1 is characterized in that: In step S1, rainfall data and its synergistic factor data of multiple time steps are used as model inputs. The synergistic factors include wind field, temperature, humidity, and air pressure variables, and ERA5 data is used as the true value source to construct a time series dataset for model training and prediction.

4. The regional extreme precipitation forecasting and warning method based on cGAN according to claim 1 is characterized in that: Step S2 specifically includes: The preprocessed multi-source meteorological data is input into the encoder of cGAN, the data is mapped to a spatial latent representation, and the KL divergence is used to regularize the latent space; The rainfall and its synergistic factor data of different spatial resolutions are uniformly processed through a differentiable interpolation layer, which consists of a convolutional layer and a pooling layer; The data is decomposed into interannual daily average data and anomaly data, and the spatiotemporal feature extraction module of the Conv-Transformer combination is used to extract features from the two types of data respectively; The inter-annual daily average features and anomaly features of multiple data are concatenated through the concat layer, the attention mechanism is used to calculate the attention of each feature, and the features are fused based on the attention scores; The fused features are subjected to spatial multi-scale feature extraction to enhance the representation capability of semantic information and fine-grained information.

5. The regional extreme precipitation forecasting and warning method based on cGAN according to claim 1 is characterized in that: Step S3 specifically includes: The enhanced multi-scale spatial features are input into the decoding module of the cGAN generator, and the potential features are mapped into the rainfall and its synergistic factor data at the next moment through layer-by-layer convolution operations; In the decoding module, the attention mechanism and feature splicing are used to further integrate multi-scale spatial feature information to predict the rainfall and its synergistic factor data at the next moment; The forecast data generated by the generator is input into the discriminator. The discriminator determines whether the generated data comes from the real data distribution by calculating the discrimination loss. This further constrains and optimizes the output of the generator, improving its ability to model real rainfall data. The generated forecast data is used as new input into the trained cGAN model to achieve multi-step prediction.

6. The regional extreme precipitation forecasting and warning method based on cGAN according to claim 1 is characterized in that: In step S4, a topological map is constructed based on the coordinates of the meteorological stations and ERA5 data grid points to represent the spatial relationship; The ERA5 rainfall data and its interpolation to the meteorological station data are used as the input of the graph neural network GNN; The observation values of meteorological stations are used as the true values, and the GNN network is trained as a post-processing module to perform spatial dependency modeling and optimization correction on the generated forecast data to improve the forecast accuracy.

7. The regional extreme precipitation forecasting and warning method based on cGAN according to claim 1 is characterized in that: In step S4, the forecast correction module is used to map the forecast data generated by cGAN to the location of the meteorological station through interpolation. The difference with the real station data is optimized and the rainfall data between adjacent grid points is adjusted in combination with the GNN spatial information transmission mechanism, thereby enhancing the spatial details of the generated regular forecast data to solve the over-smoothing problem of extreme rainfall forecasts and make the generated data more consistent with the actual surface observation rainfall data.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

9. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images