Keywords
library of scenarios, point source, puff model, bayesian statistics, probability density function
Start Date
1-7-2008 12:00 AM
Abstract
An extension of the transition of a possible source of air pollution, as a combination of measurements and inverse modelling, based on Bayesian statistics, has been proposed. The method consists of two steps. The first one is creating a library of possible scenarios, where each scenario includes meteorological parameters, parameters related to the emission and calculated concentrations at measurement points. Once the library is formed, we calculate marginal probability function for the parameter that we would like to estimate. In the most cases parameters under consideration would be those related to the emission its position, intensity or duration. The parameter that we did estimation for was the position of the possible source, creating simulated measurements that had incorporated possible measuring errors. This, the second part of calculation, is extremely efficient and fast. The simplicity of the approach and its numerical efficiency qualifies it for the problem, especially in the operational mode. Members of the library and simulated measurements were generated using a puff model.
Locating a Source of Air Pollution Using Inverse Modelling and Pre-computed Scenarios
An extension of the transition of a possible source of air pollution, as a combination of measurements and inverse modelling, based on Bayesian statistics, has been proposed. The method consists of two steps. The first one is creating a library of possible scenarios, where each scenario includes meteorological parameters, parameters related to the emission and calculated concentrations at measurement points. Once the library is formed, we calculate marginal probability function for the parameter that we would like to estimate. In the most cases parameters under consideration would be those related to the emission its position, intensity or duration. The parameter that we did estimation for was the position of the possible source, creating simulated measurements that had incorporated possible measuring errors. This, the second part of calculation, is extremely efficient and fast. The simplicity of the approach and its numerical efficiency qualifies it for the problem, especially in the operational mode. Members of the library and simulated measurements were generated using a puff model.
