In this study, the spatial change of urban green spaces in Daegu, Korea was analyzed and predicted where its urbanization is in the stage of stabilization. To do so, three kinds of analyses were performed. First, in order to analyze the spatial change of urban green spaces in Daegu from 1989 to 2009...
In this study, the spatial change of urban green spaces in Daegu, Korea was analyzed and predicted where its urbanization is in the stage of stabilization. To do so, three kinds of analyses were performed. First, in order to analyze the spatial change of urban green spaces in Daegu from 1989 to 2009, the conversion process of land cover was explored and synoptic and gradient analyses were conducted using spatial metrics. Second, using physical, social, economic, and policy factors selected, ordinary least squares(OLS) regression analysis, hot-spot analysis, logistic regression analysis and generalized estimating equation(GEE) regression analysis were performed to identify the forces driving the spatial change of urban green spaces in Daegu. Third, the CA-Markov model was applied to predict the spatial change of urban green spaces in 2020 and 2030 and to investigate the spatial patterns by synoptic and gradient analyses.
The results from this study are summarized as follows. The synoptic and gradient analyses show that urban green spaces in Daegu were gradually fragmented in size, shape, cohesion and diversity around the Dalseong-gun, Seongseo and Ansim housing development districts. Forests were most prominently fragmented in the Hwawon area and most rapidly in the Chilgok area. Grasslands were largely fragmented due to the decrease in size and cohesion in many areas and most fragmented in the Ansim area.
The GEE analysis indicates that the area and size metrics were positively correlated with the proportion of residential areas and negatively correlated with the distance to roads. In the case of the shape metrics, there was a positive relationship with the slope and the proportion of industrial areas and there were no significant variables with negative relation. The cohesion and diversity metrics were positively correlated with the slope and there were no significant variables with negative correlation.
The urban green spaces of Daegu in 2020 and 2030 were predicted using the CA-Markov model. The simulation results show that forests decreased significantly compared to the year of 2009. Through the land cover change detection, it was found that the spatial change was related to the increase of urbanized and agricultural areas. On the other hand, grasslands showed a slight increase. The synoptic and gradient analyses illustrate that the quantitative decrease of urban green spaces was observed in the region where the urban growth occurred rapidly until 2009 and they were remarkably fragmented although their complexity was lowered. In particular, Gachang-myeon in Dalseong-gun was highly fragmented. Overall, the complexity of urban green spaces was lowered but their fragmentation increased since forests almost disappeared in the areas from Doosan-dong in Suseong-gu and Bongdeok-dong in Nam-gu to Eumnae-dong and Guam-dong in Buk-gu.
The significance of this study is as follows. First of all, this study provides an empirical case study of the spatial change of urban green spaces in Daegu in the stage of stable urbanization from 1989 to 2009 considering the form and distribution along with the total amount. Second, this study made it possible to analyze the local variation of urban green spaces through the difference map by applying the moving window method as a sampling strategy for quantifying spatial metrics. Also, this study suggested the optimal window size for capturing the local variation of the spatial metrics through the sensitive analysis. Third, this study applied hot-spot analysis and GEE regression analysis to analyze the driving forces of urban green spaces. The hot-spot analysis was employed to cluster the changes of spatial metrics from 1989 to 2009 while GEE regression analysis was used to remove the spatial autocorrelation effect detected from the logistic regression analysis. Fourth, this study extends exiting literature by predicting urban green spaces in Daegu and analyzing the spatial change at the global and local levels. Finally, the spatial pattern of the fragmented urban green spaces identified by this study can be used as a base data for establishing the environment-friendly urban development strategy.
In this study, the spatial change of urban green spaces in Daegu, Korea was analyzed and predicted where its urbanization is in the stage of stabilization. To do so, three kinds of analyses were performed. First, in order to analyze the spatial change of urban green spaces in Daegu from 1989 to 2009, the conversion process of land cover was explored and synoptic and gradient analyses were conducted using spatial metrics. Second, using physical, social, economic, and policy factors selected, ordinary least squares(OLS) regression analysis, hot-spot analysis, logistic regression analysis and generalized estimating equation(GEE) regression analysis were performed to identify the forces driving the spatial change of urban green spaces in Daegu. Third, the CA-Markov model was applied to predict the spatial change of urban green spaces in 2020 and 2030 and to investigate the spatial patterns by synoptic and gradient analyses.
The results from this study are summarized as follows. The synoptic and gradient analyses show that urban green spaces in Daegu were gradually fragmented in size, shape, cohesion and diversity around the Dalseong-gun, Seongseo and Ansim housing development districts. Forests were most prominently fragmented in the Hwawon area and most rapidly in the Chilgok area. Grasslands were largely fragmented due to the decrease in size and cohesion in many areas and most fragmented in the Ansim area.
The GEE analysis indicates that the area and size metrics were positively correlated with the proportion of residential areas and negatively correlated with the distance to roads. In the case of the shape metrics, there was a positive relationship with the slope and the proportion of industrial areas and there were no significant variables with negative relation. The cohesion and diversity metrics were positively correlated with the slope and there were no significant variables with negative correlation.
The urban green spaces of Daegu in 2020 and 2030 were predicted using the CA-Markov model. The simulation results show that forests decreased significantly compared to the year of 2009. Through the land cover change detection, it was found that the spatial change was related to the increase of urbanized and agricultural areas. On the other hand, grasslands showed a slight increase. The synoptic and gradient analyses illustrate that the quantitative decrease of urban green spaces was observed in the region where the urban growth occurred rapidly until 2009 and they were remarkably fragmented although their complexity was lowered. In particular, Gachang-myeon in Dalseong-gun was highly fragmented. Overall, the complexity of urban green spaces was lowered but their fragmentation increased since forests almost disappeared in the areas from Doosan-dong in Suseong-gu and Bongdeok-dong in Nam-gu to Eumnae-dong and Guam-dong in Buk-gu.
The significance of this study is as follows. First of all, this study provides an empirical case study of the spatial change of urban green spaces in Daegu in the stage of stable urbanization from 1989 to 2009 considering the form and distribution along with the total amount. Second, this study made it possible to analyze the local variation of urban green spaces through the difference map by applying the moving window method as a sampling strategy for quantifying spatial metrics. Also, this study suggested the optimal window size for capturing the local variation of the spatial metrics through the sensitive analysis. Third, this study applied hot-spot analysis and GEE regression analysis to analyze the driving forces of urban green spaces. The hot-spot analysis was employed to cluster the changes of spatial metrics from 1989 to 2009 while GEE regression analysis was used to remove the spatial autocorrelation effect detected from the logistic regression analysis. Fourth, this study extends exiting literature by predicting urban green spaces in Daegu and analyzing the spatial change at the global and local levels. Finally, the spatial pattern of the fragmented urban green spaces identified by this study can be used as a base data for establishing the environment-friendly urban development strategy.
주제어
#도시 녹지 공간메트릭스 종관 분석 횡단축 분석 CA-Markov 핫스팟 분석 GEE
※ AI-Helper는 부적절한 답변을 할 수 있습니다.