Aristotle University Of Thessaloniki (AUTH)
Ανάπτυξη στατιστικών μεθόδων υποβιβασμού κλίμακας με εφαρμογές σε περιβαλλοντολογικά προβλήματα
Abstract
dc:descriptionThe present PhD thesis is a contribution to the study of the statistical downscaling methods. An attempt was made to research a new statistical downscaling approach, based on canonical correlation analysis, in order to construct climate change scenarios. The suggested approach focuses on two novel points: (a) the elimination of the statistical significant trend of the involved variables (predictors and predictands), by regression models using generalized least square estimators and (b) the most suitable relationship between the canonical variates, assuming that the residuals from the model-fit are serially correlated. Moreover, according to the models based on the new and the classical statistical downscaling approach, we investigate a sufficient condition that implies the suggested method is more reliable than the classical method. The statistical downscaled models are validated for a known period, using a number of distinct performance criteria. It was found that the models were able to reproduce satisfactorily the observed parameters, therefore, were used to generate climate change scenarios, in order to estimate the mean maximum summer and mean minimum winter temperatures in 2070-2100 over Greek area. From the results of the two downscaling methods, it become obvious the superiority of the proposed method.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2012
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Σκουρκέας, Αναστάσιος
- Skourkeas, Anastasios
Subjects
dc:subject × 20- Υποβιβασμός κλίμακας
- Ανάλυση κανονικών συσχετίσεων
- Ανάλυση κύριων συνιστωσών
- Τάση
- Αυτοσυσχετισμένα σφάλματα
- Γραμμική παλινδρόμηση
- Σενάρια κλιματικών αλλαγών
- Statistical downscaling
- CCA
- PCA
- Trends
- Autocorrelated errors
- Linear regression
- Climate change scenarios
- Φυσικές Επιστήμες
- Μαθηματικά
- Γεωεπιστήμες και Επιστήμες Περιβάλλοντος
- Natural Sciences
- Mathematics
- Earth and Related Environmental Sciences
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/26690