Method of and system for generating weather forecast

1. A method of generating a weather forecast, the method executable by a server, the server including a processor, the processor configured to execute a Machine Learning Algorithm (MLA), the method comprising:

receiving, by the MLA at a given period of time, an indication of meteo radar data for a given geographical region, wherein the meteo radar data being represented as a pre-determined number of sequential 3D precipitation maps having been captured before the given period of time;

receiving, by the MLA at the given period of time, satellite data for the given geographical region;

generating, by the MLA, a further sequential 3D precipitation map based on the satellite data, which follows the pre-determined number of sequential 3D precipitation maps, the further sequential 3D map being associated with a period of time after the given period of time;

generating, by the MLA, the weather forecast for the given period of time based on the pre-determined number of sequential 3D precipitation maps and the further sequential 3D precipitation map.

2. The method of claim 1 , wherein the weather forecast is a nowcasting weather prediction.

3. The method of claim 2 , wherein the nowcasting weather prediction is associated with an ultra-short prediction time span.

4. The method of claim 3 , wherein the ultra-short prediction time span is within two hours of the given period of time.

5. The method of claim 1 , wherein the generating the weather forecast further comprises generating the weather forecast for a future period of time after the given period of time.

6. The method of claim 3 , wherein the generation of the further sequential 3D map comprises applying an optical flow technology.

7. The method of claim 6 , the applying the optical flow technology is to generate at least one of:

future 3D precipitation maps based on the pre-determined number of sequential 3D precipitation maps;

future satellite data based on the satellite data.

8. The method of claim 1 , wherein the generating by the MLA the weather forecast for the given period of time is for a prediction geographical region, which is larger than the given geographical region.

9. The method of claim 1 , wherein the MLA comprises a first MLA and a second MLA.

10. The method of claim 8 , wherein the first MLA is configured to predict the further sequential 3D map based on the pre-determined number of sequential 3D precipitation maps.

11. The method of claim 10 , wherein the predetermined number is six.

12. The method of claim 10 , wherein the first MLA is a Neural Network (NN).

13. The method of claim 10 , wherein the second MLA is configured to predict the further sequential 3D precipitation map based on the satellite data.

14. The method of claim 13 , wherein the second MLA is a Neural Network (NN).

15. The method of claim 14 , wherein the Neural Network (NN) comprises a U-net NN.

16. The method of claim 13 , wherein the method further comprises training the second MLA, wherein the second MLA is trained based on a training set, the training set including a training object and a training label; and wherein:

the training object comprises the satellite data;

the training label comprises a prediction output of the first MLA, the prediction output being a further sequential 3D map.

17. The method of claim 16 , wherein the satellite data comprises:

a plurality of layers of photos made by the satellite;

a layer of topographical information;

a layer that represents the high of sun relative to the horizon;

a plurality of layers of meteo data collected by a Global Forcast System.

18. A method of generating a weather forecast, the method executable by a server, the server including a processor, the processor configured to execute a Machine Learning Algorithm (MLA), the method comprising:

receiving, by the MLA at a given period of time, an indication of meteo radar data for a given geographical region, wherein the meteo radar data having been captured before the given period of time;

receiving, by the MLA at the given period of time, satellite data for the given geographical region, the meteo radar data being represented as a pre-determined number of sequential 3D precipitation maps having been captured before the given period of time;

generating, by the MLA, an enrichment data based on the satellite data, the enrichment data being a further sequential 3D precipitation map for at least one of:

a second geographical region being at least partially different from the given geographical region; and

a period of time after the given period of time;

generating, by the MLA, the weather forecast based on the meteo radar data and the enrichment data.

19. A method of generating a weather forecast, the method executable by a server, the server including a processor, the processor configured to execute a first Machine Learning Algorithm (MLA), the method comprising:

training the first MLA by:

receiving, by a second MLA at a training period of time, an indication of meteo radar data for a training geographical region, the meteo radar data being represented as a pre-determined number of sequential 3D precipitation maps having been captured before the given period of time;

generating, by the second MLA, a prediction of a further sequential 3D precipitation map, which follows the pre-determined number of sequential 3D precipitation maps, the further sequential 3D map being associated with a period of time after the training period of time;

receiving, by the first MLA at the given period of time, satellite data for the training geographical region;

training the first MLA to generate a prediction of a 3D precipitation map based on the satellite data by providing to the first MLA a training set including:

the satellite data as a training object;

the prediction of the further sequential 3D precipitation map made by the second MLA as a target;

receiving, by the first MLA at a given period of time, satellite data for a given geographical region;

based on the satellite data, generating by the first MLA, a 3D precipitation map for the given geographical region;

based on the 3D precipitation map, generating by the first MLA the weather forecast for the given period of time for the given geographical region.