CityU research provides solution to visualize typhoon trajectory and intensity from the West Pacific and South China Sea using multi-sourced data

Introduction

Interview with the research team

Q1. How does this technology integrate with smart city management systems? For example, how can the visualization of typhoon trajectory and intensity help government departments optimize cross-department collaboration and risk planning in decisions related to traffic management, shelter resource allocation or protection for key infrastructures?

Ans: The typhoon visualization system can be embedded into the smart city management system and integrated with weather forecasts and other real-time data, providing a precise and efficient support for decision-making.

In terms of traffic management, the system integrates multi-sourced heterogeneous data from buoys, weather stations and remote sensing observations, along with real-time road conditions and traffic flow information. It generates safety emergency routes under controllable risks and optimizes signal control strategies, enhancing the collaborative scheduling efficiency of traffic and emergency management departments.

In terms of key infrastructure protection, the system combines high-precision meteorological data with the distribution of key facilities to implement tiered early warnings and differentiated protective measures, enhancing collaborative assessment and scientific response of various departments before disasters occur. Through the integration of multiple data and visualization applications, the system provides robust support and practical pathways for smart cities to respond quickly and effectively to extreme weather events, ensuring comprehensive disaster prevention and management.

Ans: This system targets high-risk typhoon areas in Hainan (such as Wenchang city and Qionghai city). Using the multi-sourced heterogeneous data integration platform, it puts together meteorological observations and satellite remote sensing data to build a spatio-temporal integrated database. Utilizing WebGIS technology, the platform enables dynamic forecasting of typhoon trajectory and visual representation of risk areas. It presents complex meteorological data in graphical, interactive interfaces, enhancing information retrieval efficiency and user comprehension. This provides intuitive, real-time and actionable early alert tools for disaster prevention and reduction.

Q3. What were the challenges, difficulties and solutions?

Ans: During the project, we encountered three main challenges:

1. Insufficient complete sets of data, there was a scarcity of high resolution spatio-temporal observational data, especially for extreme weather events;
2. Heterogeneity of multi-sourced data: Integrating data from different sources (land-based weather stations, sea-based buoys and satellite data) was complicated due to differences in format, accuracy and temporal alignment;
3. High uncertainty in predictions: The uncertainty in forecasting typhoon trajectory and intensity placed higher demands on the stability and generalization capabilities of our models.

To address these issues, we implemented the following measures:

1. Deployed buoys and weather stations in key areas to enhance the acquisition of observational data;
2. Established a unified data preprocessing and standardization framework to improve the efficiency of integrating heterogeneous data sources;
3. Experimented with the latest time series deep learning models to enhance predictive accuracy under complex weather conditions during modelling.