arXiv:2512.04385v1 Announce Type: new 
Abstract: Fine-grained air pollution forecasting is crucial for urban management and the development of healthy buildings. Deploying portable sensors on mobile platforms such as cars and buses offers a low-cost, easy-to-maintain, and wide-coverage data collection solution. However, due to the random and uncontrollable movement patterns of these non-dedicated mobile platforms, the resulting sensor data are often incomplete and temporally inconsistent. By exploring potential training patterns in the reverse process of diffusion models, we propose Spatio-Temporal Physics-Informed Diffusion Models (STeP-Diff). STeP-Diff leverages DeepONet to model the spatial sequence of measurements along with a PDE-informed diffusion model to forecast the spatio-temporal field from incomplete and time-varying data. Through a PDE-constrained regularization framework, the denoising process asymptotically converges to the convection-diffusion dynamics, ensuring that predictions are both grounded in real-world measurements and aligned with the fundamental physics governing pollution dispersion. To assess the performance of the system, we deployed 59 self-designed portable sensing devices in two cities, operating for 14 days to collect air pollution data. Compared to the second-best performing algorithm, our model achieved improvements of up to 89.12% in MAE, 82.30% in RMSE, and 25.00% in MAPE, with extensive evaluations demonstrating that STeP-Diff effectively captures the spatio-temporal dependencies in air pollution fields.

تم اقتراح إطار عمل جديد يسمى STeP-Diff لتحسين توقعات تلوث الهواء الدقيق باستخدام منصات متنقلة مزودة بأجهزة استشعار. يتناول هذا النموذج التحديات التي تطرحها البيانات غير المكتملة وغير المتسقة زمنياً من المنصات المتنقلة غير المخصصة من خلال الاستفادة من DeepONet ونموذج انتشار مستند إلى PDE للتنبؤ بمجالات التلوث الزمكانية.

Se ha propuesto un nuevo marco llamado STeP-Diff para mejorar la previsión de la contaminación del aire de grano fino utilizando plataformas móviles equipadas con sensores. Este modelo aborda los desafíos planteados por datos incompletos e inconsistentes temporalmente de plataformas móviles no dedicadas al aprovechar DeepONet y un modelo de difusión informado por PDE para predecir campos de contaminación espaciotemporales.

Un nouveau cadre appelé STeP-Diff a été proposé pour améliorer les prévisions de pollution de l'air à grain fin en utilisant des plateformes mobiles équipées de capteurs. Ce modèle répond aux défis posés par des données incomplètes et temporellement incohérentes provenant de plateformes mobiles non dédiées en s'appuyant sur DeepONet et un modèle de diffusion informé par des PDE pour prédire les champs de pollution spatio-temporels.

A new framework called STeP-Diff has been proposed to enhance fine-grained air pollution forecasting using mobile platforms equipped with sensors. This model addresses the challenges posed by incomplete and temporally inconsistent data from non-dedicated mobile platforms by leveraging DeepONet and a PDE-informed diffusion model to predict spatio-temporal pollution fields.

STeP-Diff: Spatio-Temporal Physics-Informed Diffusion Models for Mobile Fine-Grained Pollution Forecasting

Was this article worth reading? Share it

LucidQuery AI

UHoo

Dyad

Bypass

Attentive AI

Deptho.ai

Ready to build your own newsroom?