Radar echo extrapolation method based on dynamic convolution neural network

Abstract

The invention discloses a radar echo extrapolation method based on a dynamic convolution neural network. The method comprises a step of offline convolutional neural network training which comprises the steps of carrying out data preprocessing on a given training image set to obtain a training sample set, initializing a dynamic convolution neural network model, training a dynamic convolution neural network by using the training sample set, calculating an output value through network forward propagation, and updating network parameters through backward propagation such that the dynamic convolution neural network converges. The method also comprises a step of online radar echo extrapolation which comprises the steps of converting a test image set into a test sample set through data preprocessing, testing the trained dynamic convolution neural network by using the test sample set, and carrying out convolution of a laster radar echo image inputted into an image sequence and a probability vector obtained in the network forward propagation to obtain a predicted radar echo extrapolation image.

Description

technical field [0001] The invention belongs to the technical field of ground meteorological observation in atmospheric detection, and in particular relates to a radar echo extrapolation method based on a dynamic convolutional neural network. Background technique [0002] Nowcasting mainly refers to weather forecasting with a high temporal and spatial resolution of 0 to 3 hours, and the main forecasting objects include heavy precipitation, strong wind, hail and other disastrous weather. At present, many forecasting systems use numerical forecasting models, but due to the spin-up delay (spin-up) in numerical forecasting, their short-term nowcasting capabilities are limited. The new generation of Doppler weather radar has high sensitivity and resolution, the spatial resolution of its data can reach 200-1000m, and the time resolution can reach 2-15min. In addition, Doppler weather radar also has a reasonable working mode, comprehensive status monitoring and fault alarm, advanc...

AI Technical Summary

Problems solved by technology

[0003] The traditional radar echo extrapolation methods are the centroid tracking method and the tracking radar echoes by correlation (TREC) method based on the maximum correlation coefficient, but the traditional methods have certain deficiencies. Strong and small-scale storm cells are unreliable for forecasting large-scale precipitation; TREC generally regards the echo as linearly changing, but in reality, the echo changes are more complex, and this method is vulnerable to vector field disordered vector interference

In addition, existing methods have a low utilization rate of radar data, while historical radar data contain important features of changes in local weather systems and have high research value

Images