Project description/goals

This project aims to carry out research on the characteristics of the temporal and spatial distribution of sudden heavy precipitation, and to develop research on the key technologies of 0-2 hour heavy precipitation nowcasting and early warning based on artificial intelligence technology.

Importance/impact, challenges/pain points

The nowcasting business of sudden heavy precipitation is an urgent need for disaster prevention, mitigation and relief, major social activities, and refined weather forecasts. Especially in severe weather forecasts such as sudden heavy precipitation, short-term nowcasting and early warning have become prevention measures. The most important technical factor in disaster mitigation and disaster relief plays an irreplaceable role in improving the effect of disaster prevention, mitigation and disaster relief.

Solution description

We will construct the complete radar echo image data of the rainfall in Shenzhen in the past 5 years, and put forward a unified evaluation standard on this basis.

We plan to read a lot of related literature, reproduce and improve the algorithm.

Key contribution/commercial implication

It has important significance and broad application prospects to improve the accuracy of forecasting and early warning of sudden heavy precipitation weather in Shenzhen and surrounding areas, and to reduce disaster losses.

Next steps

Rain radar echo image is currently an important reference index used for precipitation prediction and radar image prediction is similar to video prediction method. 

Based on the existing video prediction research work, we are cooperating with Shenzhen Meteorological Bureau on video prediction algorithm research to improve the accuracy of radar echo image prediction.    

Collaborators/partners

Shenzhen Meteorological Bureau.

Team/contributors

Xiaoguang Han, Shuliang Ning, Mengcheng Lan, Hairui Zhu, etc.