Quantification of NO2 from different sectors of economic activity remains critical to monitor air pollutants with the rapid development of infrastructures and the rapid industrialization of emerging economies. This study aims to estimate NO2 emissions using satellite NO2 measurements from TROPOMI from different activity sectors over northern Egypt using a full-chemistry atmospheric transport model (WRF-Chem, including aerosol chemistry) and a non-linear Bayesian method. Our top-down approach was carried out for two months, January and July 2022, to analyse the seasonal variations in NO2 emissions over the studied region. The major source of uncertainties in our top-down emission estimates is due to missing sources in the prior emissions (fossil fuel inventory),  lacking more frequent updates and sub-monthly information. We also explore other sources of errors, such as the level of uncertainty used for calculating error covariance factors, the estimation of biogenic emissions, the selection of a quality assurance filter of TROPOMI NO2 and the use of regularization parameters. Non-linearities are included in our optimization algorithm by performing sensitivity experiments with the direct chemistry model (error propagation) to establish daily relationships between NO2 fluxes and concentrations. Our Bayesian inversion produces optimized values over sub-regions of Egypt and for specific sectors. We defined several subregions and assessed the regional NO2 emissions. The total estimated NO2 emissions from the studied region are about 45 Kt for the month of January 2022 and 32 Kt for the month of July 2022. The results were compared to a previous study using the flux-divergence method, showing a fair agreement for the month of winter (R2=0.67) but disagree in terms of magnitude during summer. We discuss the potential causes for the observed mismatch, which is possibly due to the extreme climate of the region, the availability of satellite observations, and large seasonal variations in the lifetime of NO2.