Spatiotemporal Predictions of Toxic Urban Plumes Using Deep Learning

TL;DR

This paper introduces ST-GasNet, a deep learning model that rapidly predicts the spread of toxic urban plumes using limited simulation data, providing accurate spatiotemporal forecasts crucial for emergency response.

Contribution

The paper presents a novel deep learning approach inspired by fluid dynamics equations that efficiently predicts toxic plume dispersion in urban environments, outperforming traditional models in speed and accuracy.

Findings

ST-GasNet achieves at least 90% accuracy in predictions.

The model accurately captures late-time plume evolution.

Incorporating wind boundary conditions improves prediction quality.

Abstract

Industrial accidents, chemical spills, and structural fires can release large amounts of harmful materials that disperse into urban atmospheres and impact populated areas. Computer models are typically used to predict the transport of toxic plumes by solving fluid dynamical equations. However, these models can be computationally expensive due to the need for many grid cells to simulate turbulent flow and resolve individual buildings and streets. In emergency response situations, alternative methods are needed that can run quickly and adequately capture important spatiotemporal features. Here, we present a novel deep learning model called ST-GasNet that was inspired by the mathematical equations that govern the behavior of plumes as they disperse through the atmosphere. ST-GasNet learns the spatiotemporal dependencies from a limited set of temporal sequences of ground-level toxic urban plumes generated by a high-resolution large eddy simulation model. On independent sequences, ST-GasNet accurately predicts the late-time spatiotemporal evolution, given the early-time behavior as an input, even for cases when a building splits a large plume into smaller plumes. By incorporating large-scale wind boundary condition information, ST-GasNet achieves a prediction accuracy of at least 90% on test data for the entire prediction period.