Eventos Anais de eventos
ENCIT 2022
19th Brazilian Congress of Thermal Sciences and Engineering
Estimation of the Deposition Rates of Atmospheric Pollutants Using Particle Swarm Optimization
Submission Author:
Roseane Albani , RJ
Co-Authors:
Roseane Albani, Vinicius Albani, Alfredo Gamboa, Davidson Moreira, Antônio Silva Neto
Presenter: Roseane Albani
doi://10.26678/ABCM.ENCIT2022.CIT22-0166
Abstract
The atmosphere continuously receives gases and particulate matter from different sources, being posteriorly removed mainly due to dry or wet deposition mechanisms. The dry deposition remotion is a major environmental issue since the gases and particulate matter falling over the earth's surface may react with other species, producing potentially harmful compounds which may threaten the underlying ecosystems in many ways. Investigating the geographical distribution and magnitude of the deposition of atmospheric pollutants is essential to decide which regions will be affected by their damaging effects, allowing the establishment of pollutant control measurements. The determination of the deposition rates can be performed using an approach based on inverse problem modelling. That includes the definition of the forward problem, which is a mathematical description of the dispersion processes related, commonly represented by an advection-diffusion partial differential equation (PDE). Also, it is necessary to provide observational data and define a technique to solve the inverse problem. Deposition rates come up in one of the terms of the advection-diffusion equation or as a flux boundary condition. However, by the nonlinear nature of the estimation process, the adjoint advection-diffusion PDE is not suitable to model the forward problem implying a significant increase in the computational cost of the inverse problem solution. In this case, the forward problem needs to be solved for each iteration of the inversion procedure. To avoid such difficulty, analytical solutions for the advection-diffusion equation, or Gaussian plume-depletion models, may be a reasonable alternative to describe the forward problem, since they are usually simpler and faster than numerical solutions. In this work, we propose a methodology based on the Gaussian plume-depletion models associated with the Particle Swarm Optimization (PSO) technique to estimate the dry deposition rates of an atmospheric pollutant. We simulate a field experiment to evaluate the proposed methodology. The PSO calculates the minimizers of the corresponding Tikhonov-type functional.
Keywords
Dry deposition, Atmospheric Pollutant Transport, Inverse problems, Particle Swarm Optimization
