Probe-Driven Visual Analytics for Diagnosing and Enhancing Physical Knowledge in Air Quality Forecasting Models

Authors

Yiming Lin (Northeast Normal University), Jinghan Bai (Northeast Normal University), Huijie Zhang (Northeast Normal University), Qiushi Xia (Northeast Normal University), Jialu Dong (Northeast Normal University)

Presentation

Session

Let's dig into the data (from China)

Time

Friday, Nov 13, 08:48 – 09:00 (US/Eastern) · session 08:00 – 09:30

Location

Hall America center

Keywords

Visual analytics, air quality forecasting, human-in-the-loop

Abstract

Air quality forecasting models increasingly incorporate physical priors, but it remains unclear whether their learned representations encode physical knowledge that can be recovered and, if so, to what extent. Answering this question is challenging because air quality forecasting is highly heterogeneous across time, space, and environmental conditions, causing the physical knowledge to become uneven or fragile in specific scenarios. We present PhysProbeVA, a probe-driven visual analytics framework for quantifying, diagnosing, and enhancing physical knowledge in air quality forecasting models. PhysProbeVA employs probing tasks around two core transport mechanisms, advection and diffusion, and uses selectivity-based measurements to estimate the recoverability of physical knowledge from frozen hidden representations. It further integrates the resulting evidence into coordinated visualizations that support progressive analysis from global assessment to regime localization, spatial diagnosis, subgroup slicing, and curriculum construction for targeted model refinement. Through a case study, a user study with 12 participants, and quantitative enhancement experiments, we show that PhysProbeVA helps users identify weak-knowledge scenarios, transform them into actionable curriculum candidates, and improve both forecasting performance and representation-level physical knowledge in the refined models.

For Practitioners

data scientists