Abstract: The goal of this work was to model the gas and aerosol chemistry in Quito, Ecuador and estimate improved emission inventories for this city; as far as we know of, this is the first time that this advanced air quality modeling is applied in Quito. For achieving this goal, the WRF-Chem (Version 4.4) model was setup for Quito, using three nested spatial domains at 32, 8 and 2 km of horizontal resolution. The emission inventories were taken from EDGAR and processed using HERMES-v3 software. The inventory for the innermost domain was spatially distributed according to: a) for industrial point sources (including energy sector), their geographical locations, b) for residential sources, population density data from the Global Human Settlement Layer, c) for traffic sources, vehicle counts at 110 measuring points throghout the city (these were also used for temporal distribution of traffic sources). Total emission inventories in the innermost domain were varied between EDGAR estimates and local estimates for Quito. In addition, biogenic emissions were estimated online by WRF-Chem (using MEGAN model). The chemical mechanism chosen was CBMZ with MADE/SORGAM aerosols (mechanism 30 in WRF-Chem). The base year of simulation was 2018, and two months were chosen for testing the emission inventories: april and december. Several different planetary boundary layer (PBL) schemes were tested: Yonsei University (YSU), Mellor-Yamada-Janjic (MYJ), Mellor-Yamada-Nakashini-Niino 2.5 (MYNN), and we also included urban canopy parameterization through the BEP module, combined with YSU, MYJ and Bougault-LaCarrere PBL options. Cumulus and aerosol radiative feedbacks were also included in the model configuration. WRF-Chem results were tested on six air quality stations across Quito metropolitan area. We found that SO2 was the most challenging gas species to model, because there was a lack of detailed information on stack source parameters, and we did not have information on industrial sources located south of Quito, which were also relevant. For CO and NOX species, reasonable results were achieved (r~0.4-0.5 and r~0.4-0.7, respectively). Ozone was the gas species with better modeled performance (r~0.7-0.8). For total PM2.5, performance was lower (r~0.4-0.5). In order to improve model performance, the local emission inventories for CO, NOx, and PM2.5 were modified by factors of 1.5, 0.75 and 3.0 approximately. On the other hand, the lack of ambient VOC and NH3 measurements prevented to improve these emission estimates. The impact of biomass emissions on ozone and PM2.5 was positive but small, and the inclusion of aerosol radiative feedback was minor, becasue of the relatively small ambient PM2.5 concentrations in Quito.

Keywords: Quito; WRF-Chem; ozone, PM2.5; emission inventories.