Hyperlocal sensing and inverse modeling reveal community impacts of urban air pollutant emissions.

Hourly, 150 m × 150 m black carbon emission maps from Bayesian inverse modeling in West Oakland, CA increase estimated emissions by ~33% by identifying underestimated and unaccounted sources.

• Why it matters: Bottom-up inventories lack fine spatiotemporal precision and often miss unpermitted or informally operated urban emission sources, limiting targeted mitigation in overburdened communities.
• What they did: Dense mobile and fixed-site measurements were combined in a Bayesian inverse modeling framework to infer hourly, hyperlocal BC emissions and assess sensitivity with ~10 monitors.
• The result: Neighborhood and port-related sources’ contribution to population-weighted exposures quadrupled from 5 to 20%, disproportionately affecting low-income neighborhoods near freight corridors, and similar performance held with reduced monitoring.