Aristotle University Of Thessaloniki (AUTH)
Ανάπτυξη αριθμητικών μοντέλων και εφαρμογή τους στη μελέτη ποιότητας του αέρα
Abstract
dc:descriptionUncertainties in the input parameters of an air quality model affect the prognostic skills of the system. However, comprehensive sensitivity analysis using traditional statistical methods like Monte Carlo are not effective for 3D models mainly due to their computational demands. Hence, in depth sensitivity analysis of 3D air quality models remains the exception and not the rule. The present work studies an alternative and computationall efficient sensitivity analysis method. The uncertainties in the input parameters are evaluated through a senses of partial derivatives. For this reason, an enhanced region of the 3D air quality model CAMx has been generated (S-CAMx) by utilising the automatic differentation tool ADIFOR. The validation of S-CAM and the interpretation of the sensitivity information is performed for domains with fundamental differences in topography, emissions and photochemistry, namely in Athens, Milan and London (AUTO - OIL II data). In every case, the S-CAMx needed one simulation in order to calculate the sensitivity indices for perturbations in all the input parameters. Additionally, the highest non-lincarity was found for perturbations at the VOC emissions.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2001
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Κιουτσιούκης, Ιωάννης
- Kioutsioukis, Ioannis
Subjects
dc:subject × 14Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/13513