Supplementary Material to "detecting Nitrogen Oxide Emissions in Qatar and Quantifying Emission Factors of Gas-Fired Power Plants - A Four-Years Study"

Annual differences in horizontal distribution of NO x emissionsThe TROPOMI data used for NO x emissions estimates were not all taken from the same version.Up to October 2021 included, we use the reprocessed S5P-PAL data (Eskes et al., 2021).This reprocessing provides a single consistent Sentinel-5P (S5P) TROPOMI NO 2 product (processed with same processorversion 2.3.1).The dataset has been generated to allow consistent data analysis (e.g.trends in COVID-19 impact on air pollution) over the period 01/05/2018-14/11/2021.For the remaining period (from November 2021 to December 2022), we use the offline stream (OFFL), with processor versions 2.3.1 from November 2021 to October 2022 included, and 2.4.0 from November 2022.As a result, the emissions displayed in Figure 6 use heterogeneous TROPOMI data.To observe the impact of these version changes on the spatial distribution of emissions, the emissions of the four years concerned are displayed in Figure S1.The main difference that is observed is in the transport term, which is lower in absolute value in 2022 in the main emitting areas.The "dipole" shape that is obtained around the main emitters, i.e.Dammam (Saudi Arabia), Manama (Bahrain), Doha (Qatar), Ras Laffan (Qatar) and the south-east of Qatar is less pronounced in 2022, without being compensated by a lower sink term.As a result, maps for 2019 to 2021 show small areas with slightly negative emissions.The study of the influence of the quality limit factor q a aims to provide an explanation for these differences.