This research consists of five chapters. Chapter 1 introduced the background and goal, scope and methodologies. Chapter 2 reviewed relevant previous researches on land use change prediction and suggested the necessity of a new methodology. Chapter 3 analyzed methodologies of spatiotemporal pattern analyses as a mean for land use change prediction and provided suggestions for a new methodology. Chapter 4 developed and verified the new method with a spatiotemporal pattern analysis to predict land use change, and predicted land use change of a city. Finally, chapter 5 suggested strategies to utilize the developed method for developing urban policy.
Chapter 1. Introduction
It is important to know the direction of market force or development pressure to cope with undesirable consequences such as unplanned development and ecosystem degradation when cities grow. There have been numerous researches to predict land use change. Nevertheless, existing methodologies for land use change prediction are scarcely practiced in real policy making since they are limited in prediction accuracy. Most land use change prediction methods rely on the past trend but the pattern of land use change is changeable according to urban policy or land policy. One of the aims of this research was to develop a new method with spatiotemporal pattern analysis to forecast land use change better. Another goal was to suggest strategies to make use of the new method in policy making.
To this end, we investigated previous researches on land use change prediction and spatiotemporal pattern analyses, and derived suggestions and limitations. Then, we formulated a new method by supplementing and expanding the existing methods. Finally, we developed a prototype system to verify the method.
Chapter 2. Literature review on land use change prediction methods and the necessity of a new method
There have been many efforts to improve the accuracy and efficacy of land use change prediction methods. Logit model as a statistical method, cellular automata making use of the evolution process of cells and agent-based model considering the interaction among agents have been popular. However, these methods are barely possible to predict land use change when human-intervened policy and development plan are involved. Furthermore, the results of those methods are often suffering from the accuracy of input data that are typically remotely-sensed and classified into some categories.
The research investigated the process of land use change in practice and analyzed data produced at each step. We selected land transaction data as one of best indicators telling market force and development pressure. In other words, undeveloped land such as agricultural or forest is likely to be developed when many land transactions are clustered spatially and temporally. A spatiotemporal pattern analysis method is required to identify the spatiotemporal clustering tendency.
Chapter 3 Literature review on spatiotemporal pattern analysis
There is no previous research on land use change prediction with spatiotemporal pattern analysis. However, spatiotemporal pattern analyses have been used as complimentary for confirmative spatial analysis in many domains such as epidemiology and criminology. This chapter reviewed the principles and applications, and identified the direction for a new method by deriving suggestions and limitations from the literature review. (The rest is omitted)
- 시공간패턴분석을 통한 토지이용변화 예측 및 활용방안 연구(Land use change prediction with spatiotemporal pattern analysis and strategies for urban policy)
- 김대종; 구형수
시공간패턴분석을 통한 토지이용변화 예측 및 활용방안 연구(Land use change prediction with spatiotemporal pattern analysis and strategies for urban policy)
|Series Title; No||국토연 / 2011-59|
|Subject Country||South Korea(Asia and Pacific)|
|Subject||Territorial Development < National Land Development|