Diurnal NOx emission underestimation constrained using overlapping TROPOMI swaths

Abstract

Existing remote sensing techniques provide insights into the global-scale daily distribution of tropospheric nitrogen dioxide (NO₂) vertical column densities (VCDs). However, they struggle to quantify the diurnal variation of nitrogen oxide (NO_x) emissions, crucial for understanding air quality and atmospheric environment. We present a novel mass-conserving inversion method and sampling strategy that uses the overlap between adjacent swath edges of polar orbiting satellite Sentinel-5P/TROPOMI to quantify five years (May 2018 to April 2023) of global diurnal NO_x emissions over upper mid-latitude regions. We account for first-order thermodynamics, chemical decay, and advective and pressure induced transport, allowing a flexible, rapid, and reliable alternative to complex models. By sampling two daytime NO₂ VCDs across consecutive orbits approximately 100 min apart, emissions rates are computed for both the 100-min window around midday as well as the 24-h daily cycle. Robust climatological estimates and uncertainty quantification are enabled over major urban areas like New York, Benelux, North China Plain, and Almaty. The estimated diurnal emissions variation show that 100-min emissions are 21 %–105 % higher than 24-h averages, which reveals differences in chemical lifetime from different energy sources, unique spatial-temporal weekend effects over China, USA, and Europe, and see-saw of emissions underestimation over cleaner regions and improvement over polluted regions. The results uncover significant new findings impacting atmospheric chemistry, remote sensing retrieval, environmental policy, and climate science respectively.