Predicting weather radar images by building a first machine learning model to generate first predictive radar images based upon input weather forecast data, and a second machine learning model to generate second predictive radar images based upon historical radar images and the first predictive radar images. Further by generating enhanced predictive radar images by providing the first machine learning model weather forecast data for a location and time and providing the second machine learning model with historical radar images for the location and an output of the first machine learning model.

A precipitation nowcasting method using dynamic motion vectors includes calculating a multiscale motion vector of a radar precipitation motion vector and a numerical model precipitation motion vector by changing spatial scale of cross correlation analysis for reach of a preset time interval from precipitation data obtained through a dual polarization radar, calculating a dynamic motion vector by merging the radar precipitation motion vector and the numerical model precipitation motion vector, generating a precipitation development and dissipation map through precipitation tracking and matching using the dynamic motion vector, and outputting a precipitation forecast field for each forecast time by applying the dynamic motion vector and the precipitation development and dissipation map to Lagrangian backward extrapolation. Accordingly, it is possible to achieve realistic precipitation forecast field simulation.

Predicting weather radar images by building a first machine learning model to generate first predictive radar images based upon input weather forecast data, and a second machine learning model to generate second predictive radar images based upon historical radar images and the first predictive radar images. Further by generating enhanced predictive radar images by providing the first machine learning model weather forecast data for a location and time and providing the second machine learning model with historical radar images for the location and an output of the first machine learning model.

A system and method for applying an adaptive adjustment or taper to an electronically scanned array (ESA) weather radar based on feedback from the weather radar. To minimize ground clutter and enable the ESA to display hazardous weather phenomena, the system adaptively adjusts amplitude and phase of ESA elements to adjust the far field pattern shape and sidelobes to maintain a desirable signal to clutter ratio. The system identifies ground clutter as a strong ground return over several azimuths depending on the radar beamwidth. Once the system IDs the ground clutter, it adaptively adjusts on receive for for the upcoming azimuths. The system selectively suppresses sidelobe echoes while maintaining the signal to noise (SNR) for weather targets. The system adaptively adjusts in real time as well as adjusting using precomputed historically accurate tapers stored in memory.

The invention concerns processing which implements the following steps to estimate polarimetric parameters: In the cross power spectrum, identifying signal-containing lines and lines only containing noise; From the power spectra of each channel and from the cross power spectrum, deleting lines at the frequencies identified as only containing noise in the cross power spectrum; Calculating polarimetric parameters as a function of the power spectra thus corrected. Application to weather radars and other types of bipolar radars having coherent reception.

The present invention is a method and system for generating an area of interest for unmanned aerial vehicle (UAV) missions. Using radar and weather data, a mission area may be generated for flights which will maximize efficiency by pre-generating flight paths based on atmospheric and other data. The UAV may include artificial intelligence (AI) capabilities for processing imaging and other sensed data. Post-processing of the data may include additional AI training and processing.

A precipitation nowcasting method using dynamic motion vectors includes calculating a multiscale motion vector of a radar precipitation motion vector and a numerical model precipitation motion vector by changing spatial scale of cross correlation analysis for reach of a preset time interval from precipitation data obtained through a dual polarization radar, calculating a dynamic motion vector by merging the radar precipitation motion vector and the numerical model precipitation motion vector, generating a precipitation development and dissipation map through precipitation tracking and matching using the dynamic motion vector, and outputting a precipitation forecast field for each forecast time by applying the dynamic motion vector and the precipitation development and dissipation map to Lagrangian backward extrapolation. Accordingly, it is possible to achieve realistic precipitation forecast field simulation.

According to one embodiment, there is provided a weather radar apparatus including means for generating observation data related to a weather phenomenon by processing a radar signal received by an antenna, and information processing means for processing the observation data. The information processing means includes means for executing recognition processing to recognize a rain area which arises as the weather phenomenon, based on the observation data, means for generating threat information for calculating a degree of threat of severe rain to a target point, based on a recognition result of the recognition processing, and means for specifying a predetermined function for calculating the degree of threat, applying the threat information as a parameter to the function, and calculating the degree of threat.

Disclosed in some examples are methods, systems, devices, and machine-readable media for 3D radar weather data rendering techniques. A computer-implemented method for 3D radar weather data rendering includes retrieving weather data from a weather radar. Gridded data is generated based on the weather data. The gridded data includes a uniform grid of cubes, where each of the cubes is associated with at least one weather parameter value of a plurality of weather parameter values corresponding to the weather data. A triangular mesh for a data grouping within the gridded data is extracted. An object file including vertices and faces associated with the triangular mesh is generated. The object file is communicated to a three-dimensional (3D) visualization system to present a 3D rendering of the object file.

A system and method for applying an adaptive adjustment or taper to an electronically scanned array (ESA) weather radar based on feedback from the weather radar. To minimize ground clutter and enable the ESA to display hazardous weather phenomena, the system adaptively adjusts amplitude and phase of ESA elements to adjust the far field pattern shape and sidelobes to maintain a desirable signal to clutter ratio. The system identifies ground clutter as a strong ground return over several azimuths depending on the radar beamwidth. Once the system IDs the ground clutter, it adaptively adjusts on receive for for the upcoming azimuths. The system selectively suppresses sidelobe echoes while maintaining the signal to noise (SNR) for weather targets. The system adaptively adjusts in real time as well as adjusting using precomputed historically accurate tapers stored in memory.