Delineating Enumeration Areas for Türkiye - A Trial On Ankara

Abstract

This thesis develops a rule-based and GIS-supported workflow for the delineation of Enumeration Areas (EAs) in Türkiye and evaluates it through the case of Ankara. The study focuses on the lack of an EA system in Türkiye that could improve the technical processes of census and sample surveys and support field implementation planning. The literature on EA production was reviewed, and the application was carried out in ArcGIS Pro and R. The number of residential units was used as the target variable, while physical barriers were treated as constraining elements. EA production was based on barrier-sensitive neighbourhood definition, urbanisation-sensitive targets, rule-based growing, and split-merge mechanisms. Based on the findings, a final workflow developed in R was proposed, together with a Shiny-based interface supporting its practical use. The pilot implementation at the neighbourhood scale showed that the workflow is applicable and that some areas may remain structurally indivisible under strict constraints. At the district scale, the workflow was applied to all neighbourhoods in Çankaya under a reference scenario and was compared with alternative scenarios that varied in terms of allowing zero-target units and splitting policies. The findings indicate that the method was able to establish an EA framework at the district level and to produce reproducible spatial and tabular outputs. Full compliance with the targeted value range, however, could not be achieved in all cases because of intense urban heterogeneity, deficiencies in barrier data, and rules designed to preserve building integrity. The main contribution of the study is the development of a transparent EA production workflow suitable for pilot implementation and institutional adaptation in Türkiye. The study shows that the use of EAs in statistical production can enable time- and cost-efficient field operations and support the production of statistics with stronger representativeness. In this respect, the adoption of EA systems in Türkiye should be regarded as a strategic investment for improving the effectiveness and quality of national statistical production.