Analysis of the Spatial and Temporal Variability of Gases and Particulate Material Concentrations in the São Paulo Metropolitan Region and its Transport to other regions using Satellite Images and Numerical Modeling

• Autor(a) principal: Benavente, Noelia Rojas
• Data de Publicação: 2022
• Tipo de documento: Tese
• Idioma: eng
• Título da fonte: Biblioteca Digital de Teses e Dissertações da USP
• Texto Completo: https://www.teses.usp.br/teses/disponiveis/14/14133/tde-19012023-125823/
• topic: AERONET; Air pollution; Air Quality station; Poluição do ar; modelo WRF-Chem; Estação de Qualidade do Ar; AERONET; produtos de satélite; satellite products; WRF-Chem model

We analyzed the spatial and temporal variability of PM2.5, NOx (NO + NO2), CO and O3 concentrations, and AOD values over southeastern Brazil, using ground-based and atmospheric column data, simulations of the WRF-Chem model version 3.9.1. Sixty air quality stations, four AERONET stations, three satellite products and four different simulation periods (June 2017, July 2018, and June and August 2019) were considered to characterize the air quality over this region. Air pollutant concentrations from a mobile station were also included to evaluate the impact of the regional transport of air pollutants generated at the Metropolitan Area of São Paulo (MASP) to neighboring areas. Input emissions comprise vehicular emissions derived from a bottom-up emission model, and on-line calculations of biogenic and fire emission rates. To determine a model configuration that is suitable for representing atmospheric photo-chemistry, first we analyzed the models ability to reproduce the meteorological conditions. Good agreement between model simulations and observations for temperature (T2m) and relative humidity (RH2m) at two meters height was found, with R > 0.85 in most periods. Underestimations occurred for the daily accumulated precipitation due to the limitations of the cloud microphysics scheme or cumulus parameterization. The model showed deficiencies to represent well the local precipitation, although it was able to reproduce precipitation patterns on a regional scale. Model predictions of PM2.5, NOx, CO and O3 agreed well with observed temporal variations and trends, with model-observation discrepancies, mainly due to uncertainties in the emission inventories. Also, simulations of NO2, CO and AOD were compared with the satellite products of OMI, MOPITT and MODIS, respectively, obtaining, in general, moderate to high correlation coefficients (0.56 < R < 0.82). The differences between the model and the satellite are related to uncertainties of the emission inventories and aero-sol model simplifications (for AOD analysis). On the other hand, the temporal evolution of CO and PM2.5 showed a strong dependence on wind speed (WS10m) and direction (WD10m) measured at 10 meters height throughout the study periods. The characteristic O3 diurnal patterns observed over urban areas, but mainly over rural, where the O3 levels were even higher due to the contributions of O3 precursors from regional transport from MASP and upwind regions in between, were well represented by the model. Both the air quality station and the WRF-Chem model showed high O3 concentrations in Botucatu (> 115 gm3), higher than that registered in the MASP ( 90 gm3) and reaching higher O3 levels than the habitual. These high O3 concentrations measured at the mobile station indicate the need for more active air quality monitoring over inland regions in southeastern Brazil where millions of people live in. This work represents a first effort combining different tools: ground-based observations, remote sensing and numerical modeling,to improve understanding of how pollutants emissions impact the air quality in the MASP and surrounding areas.