You need JavaScript to view this

Uncertainty estimation and global forecasting with a chemistry-transport model - application to the numerical simulation of air quality; Estimation de l'incertitude et prevision d'ensemble avec un modele de chimie transport - Application a la simulation numerique de la qualite de l'air

Abstract

The aim of this work is the evaluation of the quality of a chemistry-transport model, not by a classical comparison with observations, but by the estimation of its uncertainties due to the input data, to the model formulation and to the numerical approximations. The study of these 3 sources of uncertainty is carried out with Monte Carlo simulations, with multi-model simulations and with comparisons between numerical schemes, respectively. A high uncertainty is shown for ozone concentrations. To overcome the uncertainty-related limitations, a strategy consists in using the overall forecasting. By combining several models (up to 48) on the basis of past observations, forecasts can be significantly improved. This work has been also the occasion of developing an innovative modeling system, named Polyphemus. (J.S.)
Authors:
Publication Date:
Dec 15, 2005
Product Type:
Thesis/Dissertation
Report Number:
FRNC-TH-6681
Reference Number:
RN07000066; TVI: 0610
Resource Relation:
Other Information: TH: These mathematiques et informatique; [110 refs.]
Subject:
54 ENVIRONMENTAL SCIENCES; AIR QUALITY; ATMOSPHERIC CHEMISTRY; GENERAL CIRCULATION MODELS; OZONE; PHOTOCHEMICAL REACTIONS; METEOROLOGY; POLLUTANTS; LAND USE; COMPUTERIZED SIMULATION; ADVECTION; DIFFUSION; MONTE CARLO METHOD; FORECASTING; DATA COVARIANCES; DAILY VARIATIONS; PROBABILISTIC ESTIMATION
OSTI ID:
20802726
Research Organizations:
Ecole Nationale des Ponts et Chaussees (ENPC), 75 - Paris (France)
Country of Origin:
France
Language:
French
Other Identifying Numbers:
TRN: FR0603374
Availability:
Commercial reproduction prohibited; OSTI as DE20802726
Submitting Site:
FR
Size:
191 pages
Announcement Date:
Dec 22, 2006

Citation Formats

Mallet, V. Uncertainty estimation and global forecasting with a chemistry-transport model - application to the numerical simulation of air quality; Estimation de l'incertitude et prevision d'ensemble avec un modele de chimie transport - Application a la simulation numerique de la qualite de l'air. France: N. p., 2005. Web.
Mallet, V. Uncertainty estimation and global forecasting with a chemistry-transport model - application to the numerical simulation of air quality; Estimation de l'incertitude et prevision d'ensemble avec un modele de chimie transport - Application a la simulation numerique de la qualite de l'air. France.
Mallet, V. 2005. "Uncertainty estimation and global forecasting with a chemistry-transport model - application to the numerical simulation of air quality; Estimation de l'incertitude et prevision d'ensemble avec un modele de chimie transport - Application a la simulation numerique de la qualite de l'air." France.
@misc{etde_20802726,
title = {Uncertainty estimation and global forecasting with a chemistry-transport model - application to the numerical simulation of air quality; Estimation de l'incertitude et prevision d'ensemble avec un modele de chimie transport - Application a la simulation numerique de la qualite de l'air}
author = {Mallet, V}
abstractNote = {The aim of this work is the evaluation of the quality of a chemistry-transport model, not by a classical comparison with observations, but by the estimation of its uncertainties due to the input data, to the model formulation and to the numerical approximations. The study of these 3 sources of uncertainty is carried out with Monte Carlo simulations, with multi-model simulations and with comparisons between numerical schemes, respectively. A high uncertainty is shown for ozone concentrations. To overcome the uncertainty-related limitations, a strategy consists in using the overall forecasting. By combining several models (up to 48) on the basis of past observations, forecasts can be significantly improved. This work has been also the occasion of developing an innovative modeling system, named Polyphemus. (J.S.)}
place = {France}
year = {2005}
month = {Dec}
}