Skip to content
Xavier WANGMIND · MODELS · MOTIONAI × DISASTER SCIENCE

MIND · MODELS · MOTION

Xavier WANG

Understanding Hazard Dynamics with World Models

Developing AI methods for safer Earth systems — finding patterns in data, understanding complexity through models, and working toward a more resilient future.

Explore my researchAbout me
  • MODELS
  • DATA
  • RISK
  • A SAFER WORLD
Explore Research

The research process

RESEARCH PROCESS

  1. 01

    Observation

    Observe Earth systems through diverse data

  2. 02

    Modeling

    Represent complexity in computable models

  3. 03

    Evolution

    Explore how systems change over time

  4. 04

    Risk prediction

    Inform safer decisions and action

Research areas

RESEARCH AREAS

Explore all research

RESEARCH JOURNEY

From computer science to disaster intelligence

Education

  1. 2020 — 2024

    Computer Science

    A foundation in algorithms, systems, and software engineering

  2. 2021 — 2023

    Software and Data Systems

    Building backends, databases, and data pipelines that work in practice

  3. 2023 — 2024

    IoT and Data Analytics

    From device-level data collection and transmission to modeling and analysis

  4. 2024 — 2025

    Generative AI and Medical Video

    Modeling dynamic scientific data — from clinical text to angiography videos (JAMA Ophthalmology, 2025)

  5. 2024 — Present

    Disaster Intelligence

    Multi-hazard early warning and rainfall modeling for real-world decisions — toward world models for complex environmental systems

Research systems & engineering practice

SYSTEMS & PRACTICE

View all projects

PRECIPITATION NOWCASTING & MULTI-HAZARD EARLY WARNING

RainGod-100

R A I N G O D - 1 0 0

A foundation model for precipitation nowcasting

Combining satellite, radar, and rain-gauge observations with physical modeling and artificial intelligence to forecast rainfall up to two hours ahead, supporting early warning for flash floods, debris flows, landslides, and urban flooding.

Forecast lead time
0–2 h
Satellite · radar · gauges
Multi-source
Data meets mechanisms
AI + Physics
Explore the projectContributed to model design, training, and release

SCIENCE IN SERVICE
OF A SAFER TOMORROW

Rainfall forecast exampleTen-minute rainfall intensity over China
RainGod-100 ten-minute rainfall-intensity forecast over China on August 11, 2026
Observation inputsSatelliteWeather radarRain gauges

LoRaWAN & data analytics

LORAWAN & DATA ANALYSIS

Sensing system

From end devices to data analysis

Building low-power collection and transmission networks, with data cleaning, statistics, visualization, and exploratory modeling.

View details

Blockchain & digital wallet

BLOCKCHAIN & WALLET

Engineering

From software to on-chain interaction

Developing a blockchain and digital wallet system, with deployment and interaction verified on an Ethereum-compatible network.

View details

Multi-hazard foundation models

MULTI-HAZARD FOUNDATION MODELS

Research direction

Representing states and evolving systems

Exploring world models of environmental states, evolving dynamics, and possible futures for complex hazard processes.

Explore research

Mountain hazards & risk prediction

MOUNTAIN HAZARDS & RISK

Research direction

Understanding processes, informing decisions

Studying landslides, debris flows, and floods with remote sensing, geospatial information, and spatiotemporal observations.

Explore research

Selected publication

SELECTED PUBLICATION

View all publications
Official JAMA Ophthalmology publication webpage
JAMA Ophthalmology · Publication and generated imagery

2025JAMA OphthalmologyCo-first Author

Generation of Fundus Fluorescein Angiography Videos for Health Care Data Sharing

JAMA Ophthalmology. 2025; 143(8): 623–632 · DOI: 10.1001/jamaophthalmol.2025.1419

This work develops a text-to-video model that generates dynamic fundus fluorescein angiography videos from clinical reports, exploring privacy-preserving medical data sharing, clinical visualization, and education.

Read PublicationProject details

Data sharing

Exploring privacy-preserving sharing of high-quality medical data

Clinical visualization

Translating clinical reports into dynamic medical images

Medical education

Exploring applications in teaching, training, and public education

FROM REAL-WORLD PROBLEMS TO LASTING IMPACT

Exchange & collaboration

COLLABORATION

MIND · MODELS · MOTION

OPEN SCIENCE

More open scienceA more resilient world

Open science for
a more resilient world.

SCIENCE IN SERVICE OF A SAFER TOMORROW.

Find questions at the intersections. Advance research together.

Exploring intelligence in complex environments

Hazard evolution, environmental forecasting, and intelligent decision-making are important areas at the intersection of artificial intelligence and scientific computing.

My research focuses on combining world models, generative methods, and scientific data. I welcome researchers and engineers from all fields to exchange ideas on these questions and advance method validation and practical applications together.

  • World models
  • Generative methods
  • Scientific data
Explore my research background

Academic exchange

ACADEMIC EXCHANGE

Explore shared questions in disaster intelligence, world models, and generative methods.

Engineering practice

ENGINEERING PRACTICE

Connect research and practice, from data collection and model development to working systems.

FROM SHARED QUESTIONS TO REAL-WORLD IMPACT.