Jun ZhuPei DangYungang CaoJianbo LaiYukun GuoPing WangWeilian LiFaculty of Geosciences and Environmental Engineering,Southwest Jiaotong University,Chengdu,PR ChinaJun Zhu earned his M.S. in geodesy and survey engineering from Southwest Jiaotong University in 2003 and a Ph.D. in cartography and GIS from the Chinese Academy of Sciences in 2006. He's now a Professor at Southwest Jiaotong University's Faculty of Geosciences and Environmental Engineering,specializing in computer vision,3-D GIS and virtual geographic environments. He provided a crucial dataset for this research and offered valuable suggestions on data analysis methods,ensuring the accuracy and completeness of the study.Pei Dang obtained his B.S. in GIS from Southwest Jiaotong University in 2016 and an M.S. in survey engineering in 2022 from the same institution. He's currently a Ph.D. candidate there,researching virtual geographic environments,spatial cognition,and GeoAI. He played a key role in the literature review section,helping to determine the direction and framework of the research,and provided several important references.Yungang Cao obtained his Ph.D. in cartography and geographic information systems from the Chinese Academy of Sciences in 2006. He is currently a professor at Southwest Jiaotong University. His research includes 3D reconstruction and environmental remote sensing. He provided technical support for this research,especially in software and programming,ensuring smooth and accurate data processing.Jianbo Lai obtained his B.S. and M.S. in GIS from Northwest Normal University. He is currently a Ph.D. candidate there,researching virtual geographic environments and city planning. He wrote key code for this study.Yukun Guo obtained his B.S. and M.S. in GIS from Southwest Jiaotong University. He's currently a Ph.D. candidate there,researching virtual geographic environments and disaster management. He designed a survey questionnaire and was responsible for recruiting participants in this study.Ping Wang obtained his B.S. and M.S. in GIS from Lanzhou Jiaotong University. She's currently a Ph.D. candidate there,researching deep learning and image detection. She is responsible for analyzing and checking the data,and participated in the distribution and collection of survey questionnaires.Weilian Li obtained his Ph.D. from Southwest Jiaotong University in 2021. He is currently conducting research on 3D GIS and spatial cognition at the University of Bonn in Germany and in Shenzhen. He provided valuable field investigation data and observations during the research process,offering empirical support for the study.

Abstract:Public's rational flood mitigation behaviors depend on accurate perception of flood risks. The use of natural language for flood risk perception is an effective approach, and it is critical to ensure the accuracy and comprehensibility of the flood information provided by the system in natural language dialogues. This study presents a framework for large language model (LLM) that is constrained by flood knowledge and can interact with geographic information system (GIS), aimed at enhancing the public's perception of flood risks. We tested the performance of LLM within this framework and the results demonstrate that LLM can generate accurate information about floods under the constraints of entities and relationships in the knowledge graph, and interact with GIS to produce personalized knowledge through real-time coding. Furthermore, we conducted flood risk perception experiments on users with different cognitive levels. The results indicate that using natural language dialogue can narrow the differences brought about by cognitive levels, allowing the public to equally access knowledge related to flood events.

Towards Democratized Flood Risk Management: An Advanced AI Assistant Enabled by GPT-4 for Enhanced Interpretability and Public Engagement

Pixels and Predictions: Potential of GPT-4V in Meteorological Imagery Analysis and Forecast Communication

A flood knowledge-constrained large language model interactable with GIS: enhancing public risk perception of floods

Artificial Intelligence Methodologies Applied to Prompt Pluvial Flood Estimation and Prediction

AI Increases Global Access to Reliable Flood Forecasts

A 3D virtual geographic environment for flood representation towards risk communication

Challenges and opportunities when bringing machines onto the team: Human-AI teaming and flood evacuation decisions

FloodBrain: Flood Disaster Reporting by Web-based Retrieval Augmented Generation with an LLM

Deep learning rapid flood risk predictions for climate resilience planning

From Data to Decision: Interpretable Machine Learning for Predicting Flood Susceptibility in Gdańsk, Poland

Breaking Down the Computational Barriers to Real‐Time Urban Flood Forecasting

Unraveling the complexities of urban fluvial flood hydraulics through AI

FloodLense: A Framework for ChatGPT-based Real-time Flood Detection

The Power of Explainability in Forecast-Informed Deep Learning Models for Flood Mitigation

FloodGenome: Interpretable Machine Learning for Decoding Features Shaping Property Flood Risk Predisposition in Cities

Presenting hydrological and data driven flood simulation-prediction methods to develop a decision-making model

A novel AI-based model for real-time flooding image recognition using super-resolution generative adversarial network

Improving Community Resiliency and Emergency Response With Artificial Intelligence

Urban Flood Risk Assessment through the Integration of Natural and Human Resilience Based on Machine Learning Models

A Rapid Assessment Method for Flood Risk Mapping Integrating Aerial Point Clouds and Deep Learning

Flood Susceptibility Assessment in Urban Areas via Deep Neural Network Approach