A recent study by researchers at the Penn State College of Earth and Mineral Sciences has revealed an unexpected dynamic in methane across the American metropolis of Baltimore: emissions from natural gas leaks are not only seasonal; they also fluctuate from hour to hour, creating a daily pattern that traditional inventory methods fail to capture.
What the study reveals: detailed findings and key figures
By combining several years of satellite observations from the Sentinel‑5P mission with ground-based measurements, scientists were able to identify areas where methane concentrations consistently exceeded background levels. They found that concentrations rose markedly during the coldest months, when heating demand intensified leaks, then declined in spring before edging up again in summer, coinciding with increased energy use for air conditioning. These seasonal changes were accompanied by daily cycles, with emission peaks generally occurring during the morning and evening peak periods, corresponding to times when heating and ventilation systems were switched on.
Researchers emphasize that, although the press release provides no absolute figures, the shape of the signal suggests that urban leaks may account for several percent of the total methane emissions attributed to the metropolitan region. Although this contribution is modest on a national scale, it represents a significant share of the local carbon budget, particularly given that methane has a global warming potential approximately 28 times that of CO₂ over a 100-year horizon. Thus, every undetected leak adds to the uncertainty in forecasting the city's climate balance.
Finally, the study highlights a pronounced geographic disparity: industrial districts and older residential neighborhoods record the highest concentrations, while recently renovated areas or those served by modernized distribution networks show levels close to background. This fine-grained mapping paves the way for targeted interventions, prioritizing the oldest networks and sites where hourly variations are most pronounced.