校園土地利用之地景變遷——以國立中興大學為例

Landscape Change of Land-Use in Campus—A Case Study of National Chung Hsing University

本文研究主旨在於（1）運用地理資訊系統（GIS）建立校園屬性資料，並以此功能輔助後續研究之用；（2）利用地景結構指數探討中興大學校園的變遷過程；（3）以馬可夫鏈機率向量的基本原理與概念，配合地理資訊系統之相關技術，探討校園土地利用型之地景結構與變化，（4）配合以時間歷程為考量的半馬可夫鏈，來推測未來校園土地利用於穩定狀態下所需的分配比例。本研究由地景結構指數於地景變遷結果中，從1982~1990 年的土地利用型圖得知，校園土地利用因公共工程而有所轉換，這是將破碎的嵌塊體經過規劃，合併並整合；經由馬可夫鏈向量機率的求解後，得到1995~2001 年之間的穩定矩陣，為校園土地利用之轉置機率矩陣，其結果為土地利用型轉移到綠地、道路、建物、荒地、水體之比例分別為：41.94 %、34.72 %、21.27 %、0.52 %、0.55 %；由半馬可夫鏈的時間過程中，得到校園未來土地利用的發展比例為，綠地：54.83 %；建物：30.85 %；道路：14.06 %；荒地：0.03 %；水體：0.23 %，由此可知，校園在未來土地發展中，綠地佔較大比例，因此，未來校園將可再創造更多的綠地，建物則是以建坪的增加，以減少建地的擴展，藉此提供規劃者規劃校園時的參考依據。 The main purposes of this study are: (1) to apply GIS to establish the geo-database of campus and support other researchers; (2) to study the change process of NCHU using landscape structure indecies; (3) to study the landscape structure and change of campus land use by Markov probability vector and GIS; (4) to predict the necessary distribution of campus land use under steady state in the future using the semi-Markov chain with time interval. From this study the campus administrator could have a reference for planning land use of campus in the future. The results of campus landscape structure index from 1982 to 1990 showed that the change of land use was derived from the public constructions. The land use of campus in 1982 were shown to be organized and merged. By solving Markov chain probability vector, the stable probability vector in 1995 to 2001 is the land use stable probability vector in the campus. The result indicated: the area proportions of different land uses are green land 41.94%, buildings 34.72%, road 21.27%, wasteland 0.52% and water 0.55%. The result showed the land use of campus is almost stable, and the present state of the land use is saturated. From the time process of Semi-Markov chain, the land use proportion of future campus will be: green land 54.83%, buildings 30.85%, road 14.06%, wasteland 0.03% and water 0.23%. By this result, the green land would occupy a larger proportion of the campus area in the future. The planner should take the green land as the main sort of land use in the future.

The main purposes of this study are: (1) to apply GIS to establish the geo-database of campus and support other researchers; (2) to study the change process of NCHU using landscape structure indecies; (3) to study the landscape structure and change of campus land use by Markov probability vector and GIS; (4) to predict the necessary distribution of campus land use under steady state in the future using the semi-Markov chain with time interval. From this study the campus administrator could have a reference for planning land use of campus in the future. The results of campus landscape structure index from 1982 to 1990 showed that the change of land use was derived from the public constructions. The land use of campus in 1982 were shown to be organized and merged. By solving Markov chain probability vector, the stable probability vector in 1995 to 2001 is the land use stable probability vector in the campus. The result indicated: the area proportions of different land uses are green land 41.94%, buildings 34.72%, road 21.27%, wasteland 0.52% and water 0.55%. The result showed the land use of campus is almost stable, and the present state of the land use is saturated. From the time process of Semi-Markov chain, the land use proportion of future campus will be: green land 54.83%, buildings 30.85%, road 14.06%, wasteland 0.03% and water 0.23%. By this result, the green land would occupy a larger proportion of the campus area in the future. The planner should take the green land as the main sort of land use in the future.