Back to search

Aristotle University Of Thessaloniki (AUTH)

Δημιουργία μοντέλου πρόβλεψης μεταβολών χρήσεων γης σε αστικές και περι-αστικές περιοχές με τη χρήση των δεδομένων του κτηματολογίου

Abstract

dc:description

The present Ph.D thesis forms part of the broader research regarding the exploration and documentation of the processes that lead to land use change at an urban and peri-urban scale. The proposed methodology involves the use of cellular automata theory. The model also tries to integrate the cadastral spatial and descriptive data. A cellular automata system can be defined by the following basic parameters: a spatial reference unit, a set of attribute values that each reference unit can take, a clearly defined neighborhood surrounding each reference unit and a defined time period at the end of which the attribute values will be updated. The model considers the cadastral plot as its prime spatial reference unit. The main attribute of each reference unit is its most dominant land use as recorded in the cadastral database. The neighborhood definition that was used in the model presents two distinct differentiations from the standard C.A models. Firstly, the land uses that were present in each neighborhood definition were represented by their corresponding total percentage and not in the form of an absolute figure. The influence of the area occupied by each land use was also studied through its corresponding percentage. These two assumptions were used in order to reduce the sensitivity of the model to the different neighborhood definitions. A time period of ten years was selected for the update of the parcels’ land use. The cadastral database was used as the main source for geospatial and descriptive data. The information was complemented by additional external sources regarding the recording of various institutional restrictions and the recording of land uses that take place through short-term leases. The calibration of the model was performed using machine learning techniques and Rough Set Theory based on three distinct historical datasets regarding the years 1993, 2003 and 2013. Historical data can be extracted through the use of sequential logical queries on the descriptive cadastral database and verified through historical aerial or satellite imagery. The land use change model was implemented at the municipalities of Kordelio-Evosmos, Ampelokipoi-Menemeni and the municipal units of Oreokastro and Ionia, on the west of the city of Thessaloniki in Northern Greece. The calibration of the model resulted in 12 individual simulation classifications which were assessed with regards to their performance and accuracy. It was concluded that the use of a percentage instead of an absolute figure assisted in reducing the sensitivity of the results to the definition of the neighborhoods, as the recorded achieved accuracy difference in the results was around 1%. The application of the cross validation method improved the overall accuracy by 0.5-2%, while increasing the number of the transition rules by 30%. The classification results are particularly encouraging especially in the cases where there was an adequate number of recorded changes towards a specific land use e.g the residential land use classification presented an accuracy of 82% in the historical dataset and an overall accuracy of 97% in the entire study area.

Degree

thesis:*
Grantor dc:publisher
Aristotle University Of Thessaloniki (AUTH)
Year dc:date
2019

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Chalkidou, Sevasti
  • Χαλκίδου, Σεβαστή

Subjects

dc:subject × 12

Rights

dc:rights
Statement dc:rights
  • BY_NC_SA
Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/46830

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Chalkidou, Sevasti; Χαλκίδου, Σεβαστή. Δημιουργία μοντέλου πρόβλεψης μεταβολών χρήσεων γης σε αστικές και περι-αστικές περιοχές με τη χρήση των δεδομένων του κτηματολογίου. Aristotle University Of Thessaloniki (AUTH), 2019. http://hdl.handle.net/10442/hedi/46830