With prospective applications ranging from improving the understanding of the daily and seasonal dynamics of noise exposure to raising public awareness of the associated health effects, dynamic noise mapping in real time is one of the next milestones in environmental acoustics. The present contribution proposes a methodology for near-real-time dynamic noise mapping, enabling the generation of dynamic noise maps and the calculation of advanced noise exposure indicators, here arbitrarily established for the previous day, on the scale of large urban areas. This methodology consists in (i) collecting live traffic counts, measured using dedicated IoT sensors, (ii) calibrating a microscopic traffic simulation using these sparsely distributed traffic counts, (iii) modelling noise emission and propagation from the microscopic traffic simulation, and finally, (iv) post-processing the noise simulation output for the calculation of a wide range of exposure indicators. The applicability of the method is demonstrated on the city of Tartu, Estonia.