Aristotle University Of Thessaloniki (AUTH)
Μελέτη μοντελοποίηση και πρόβλεψη ακραίων κλιματικών παραμέτρων στη Λεκάνη της Μεσογείου
Abstract
dc:descriptionThe increasing trend of the frequency and intensity of temperature and precipitation extremes during the recent decades has substantial environmental and socioeconomic impacts. Thus, the objective of the present dissertation is to study, model and predict extreme temperature and precipitation episodes in the Mediterranean area. This is firstly achieved with the comparison of several mathematical distributions, in order to identify the most appropriate for the description of extreme precipitation, and high and low temperature episodes in Mediterranean. The selection of the distribution that can satisfactorily fit extreme events was made after a sequence of statistical tests. Additionally, as the method that is used for the calculation of the critical parameters of a distribution is an important factor for the results, three methods have been tested in order to select the most appropriate (MLE, L-moments, Bayesian). The results showed that, the GPD can characterize accurately both precipitation and temperature extremes while the GEV distribution with the Bayesian method is also appropriate especially for the greatest values. Additionally, the present dissertation, introduces, evaluates and uses a new statistical method, namely TIN-Copula method which is a combination of triangular irregular networks and copulas. The new method is proposed for the bias correction of regional climate models’ projections for extreme climate events, as well as for the simulation of extreme maximum and minimum temperatures and precipitations at areas with no available data. The concept of the TIN-Copula method starts with the division of the studied area in non-overlapping irregular triangles which are shaped using the location of the available stations. After that, using copulas, the extreme values of an unknown point included in a triangle are bias corrected or estimated. In the calculations, also the distances between the unknown point and the triangle vertices are taken into account. Finally, the evaluation of the new method proved that the ΤΙΝ-Copula is a really useful tool, both for bias correction and estimation of extreme climate episodes in Mediterranean area.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Lazoglou, Georgia
- Λάζογλου, Γεωργία
Subjects
dc:subject × 9Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/46862