University of Ioannina
Αντικειμενικός ορισμός και μεταβλητότητα των εποχών στην περιοχή της Μεσογείου
Abstract
dc:descriptionThis doctoral dissertation is an extensive attempt to objectively define the seasons for the Mediterranean region during the period 1949-2018 using multiple methods. The aim of this research effort is to create a “new” definition of the seasons, different from the conventional one, based on objective criteria. Also, possible changes in the main characteristics of the seasons (start/end dates and duration) during the 70-year study period are investigated. In the first part of this dissertation, the definition of seasons includes the study of the mean inter-annual variation of 12 meteorological parameters. The data are obtained from the NCEP/NCAR database and are daily values (12UTC) of the following parameters: precipitation rate, convective precipitation rate, temperature at 2 meters above the earth's surface, total cloud cover, zonal and meridional wind components at 10 meters above the earth's surface, 500hPa and 1000hPa geopotential heights, 500hPa and 850hPa air temperature, 850hPa specific humidity and precipitable water. First, the mean intra-annual variations of the above parameters are calculated and Principal Component Analysis (PCA) is applied, in order to reduce the large amount of data, and next k-mean Cluster Analysis (CA) is applied on the time series of the Principal Components in order to group the dates of the year with similar patterns of the above parameters. The date groups obtained can be considered as the objectively defined seasons. This method, called Meteorological Variables mean intra-annual Variation method (MVV), is applied again for the five overlapping 30-year subperiods: 1949-1978, 1959-1988, 1969-1998, 1979-2008 and 1989-2018, with the aim of investigating possible long-term changes of the seasons’ characteristics. According to the results, 4 seasons are defined for all time periods. Although these seasons correspond to the four conventional seasons, there are differences in their onset/cessation dates and duration. According to the MVV method, winter lasts about 4 months, spring and summer last a little less than 3 months and autumn lasts about 2.5 months. The most notable long-term changes in the characteristics of the seasons are: i) the decrease in the duration of winter and spring due to the delay of their onset, and ii) the extension of autumn due to the delay of its end. In the second part of the dissertation, another method of objectively defining the seasons is adopted, which is based on the definition of Weather Types (WTs). The meteorological parameters used in this method are the same as those used in the MVV method but without taking into account the two precipitation parameters. Initially, PCA and CA are applied to the daily values of the parameters of the 70-year period 1949-2018 and groups are formed whose dates have common synoptic characteristics. Each of these groups corresponds to a Weather Type (WT). Therefore, 8 WTs are defined and their characteristics are studied, including the prevailing synoptic conditions and their monthly and annual frequency distributions. Then, PCA and CA are applied again, but this time on the intra-annual variations of the WTs’ frequencies. As a result, the dates within the year that have a similar frequency distribution among the 8 WTs are classified into groups which can be considered as the objectively defined seasons. According to this method, which is called Weather Types’ Frequency method (WTF), winter and summer last about 4 months and spring and autumn last about 2 months. The most significant changes in the limits and duration of the seasons resulting from the WTF method are: i) the shortening of winter and spring in the most recent periods, related to later start dates and earlier end dates, respectively, ii) the extension of autumn and summer, due to later end dates, and earlier start and later end dates, respectively. In the third part of the thesis, the definition of seasons is based on the cyclones and their tracks which affect the Mediterranean region. The analysis is performed using Mean Sea-Level Pressure (MSLP) data with a temporal analysis of 6 hours (00, 06, 12 and 18 UTC) and high spatial resolution (1°×1°). The data are obtained from the state-of-the-art reanalysis database ERA5 and refer to the period 1950-2018. The main cyclogenesis and high-density areas are identified and are the Gulf of Genoa, the Atlas Mountain range, the Iberian Peninsula, the Cyprus region, the Black Sea and the Middle East. Then, a statistical analysis is performed using PCA and CA, in order to classify the cyclone tracks that were detected into groups and as a result 12 Cyclone tracks’ Clusters (CCs) are revealed. The main characteristics of the 12 CCs are analyzed, such as their preferred paths, their persistence throughout the year as well as the areas affected. Also, by utilizing the mean intra-annual variations of the frequencies of the 12 CCs, the Mediterranean seasons are defined. According to the results, four seasons are objectively defined which correspond to the conventional ones but present differences in their limits and duration. This method is called Cyclones Tracks’ Frequency method (CTF). In contrast with the two previous methods used, the MVV and WTF, the seasons defined by the CTF method are not continuous, which makes it difficult to distinctly define the seasons. According to the CTF method, winter lasts 4 months, summer lasts 3.5 months while spring and autumn last a little longer than 2 months. Also, the application of the CTF method on the 5 overlapping 30-year subperiods, as was followed in the previous methods, results in the definition of five seasons in some cases. Finally, a synthesis method (S method) of the above three methods, the MVV, WTF and CTF, is applied. According to this method, CA is applied to the unified matrix of the Principal Components of the three individual aforementioned methods. Therefore, the S method is a complete attempt to objectively define the Mediterranean seasons, since it takes into account the mean intra-annual variations of a variety of meteorological parameters, the prevailing weather types and the depressions affecting the region under study. According to the S method, the seasons are as follows: winter which begins on November 16 and ends on March 25 lasting more than 4 months (131 days), spring which is the period March 26 - June 11 with a duration of about 2.5 months (78 days), summer which follows until September 12 and lasts 3 months (93 days) and finally autumn (September 13 - November 15) which closes the annual cycle with a duration of 2 months (63 days). Also, the long-term changes of the characteristics of the seasons are investigated and it is found that: winter starts later and ends later in recent years and has become shorter, spring arrives and leaves earlier in recent periods and lasts a little less, summer comes earlier and ends later and therefore is longer, and finally autumn begins and ends later and its duration increases.
Degree
thesis:*- Grantor dc:publisher
- University of Ioannina
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kotsias, Georgios
- Κώτσιας, Γεώργιος
Subjects
dc:subject × 16- Ορισμός των εποχών
- Μεσογειακό κλίμα
- Φασματική ταξινόμηση
- Τύποι καιρού
- Μεσογειακές Υφέσεις
- Definition of seasons
- Mediterannean climate
- Spectral clustering
- Weather types
- Mediterannean cyclones
- Γεωεπιστήμες και Επιστήμες Περιβάλλοντος
- Φυσικές Επιστήμες
- Μετεωρολογία και Επιστήμες της ατμόσφαιρας
- Earth and Related Environmental Sciences
- Natural Sciences
- Meteorology and Atmospheric Sciences
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/52670