發(fā)布時間：2018-01-19 13:21

  本文關(guān)鍵詞： 不透水面 決策樹分類 空間分析 城市擴張 城市蔓延 景觀格局　出處：《昆明理工大學》2013年碩士論文　論文類型：學位論文

【摘要】：城市化是20世紀以來最為顯著的人類活動特征,隨著人口的快速增長以及科技的飛速進步,人類對區(qū)域土地的改造速度與需求與日俱增,城市化進程日益加劇。不透水面作為高速城市化進程中人為改造地表活動的特征產(chǎn)物,其面積的迅速增加很大程度上影響著區(qū)域生態(tài)環(huán)境變化并主導著城市景觀格局演變過程。作為城市的基質(zhì)景觀及典型地表類別,不透水面常被用于地表覆被動態(tài)監(jiān)測、城市水文模擬、非點源污染負荷估算、熱環(huán)境效應、區(qū)域氣候變化、城市擴張監(jiān)測以及城市生態(tài)系統(tǒng)評價等多個方面。 本文以昆明市為研究區(qū),在遙感、GIS以及景觀格局分析軟件等支持下,采用改進決策樹方法提取不透水面信息并應用皮爾遜卡方檢驗、香農(nóng)熵等統(tǒng)計分析模型進行空間擴張定量分析,從政策導向、地形等方面進行擴張驅(qū)動力分析并從城市生態(tài)景觀層面進行區(qū)域景觀格局分析。文章基于研究區(qū)2002、2006和2010年三期Landsat ETM+影像開展研究,選取影像的3、4、5、7波段以及圖像增強處理過程中主成分分析后的第一波段和歸一化植被指數(shù)運算后的植被信息增強波段共6個波段進行影像組合。采用基于CRUISE算法的決策樹分類方法進行研究區(qū)土地覆被類別解譯,并采用改進歸一化水體指數(shù)以及基于植被覆蓋度的不透水面密度分割圖進行掩膜處理,進一步提高解譯精度,得到研究區(qū)三個時期的土地利用分類信息。 利用空間象限法將研究區(qū)劃分為8個方向區(qū)域,引入皮爾遜卡方檢驗、香農(nóng)熵、優(yōu)勢度、擴張強度指數(shù)等分析模型從方向區(qū)域、發(fā)展階段及總體情況等三個層面進行研究區(qū)不透水面空間擴張定量分析。結(jié)果顯示,各方向區(qū)域不透水面擴張速度和擴張趨勢存在較大差異,體現(xiàn)出空間擴張的不均衡性。2002～2006年間,城市不透水面擴張受政策、規(guī)劃等因素的約束較小,發(fā)展較為自由但擴張速度相對緩慢。2006～2010年間,其增長速度有所加快,擴張優(yōu)勢度有較為明顯的提升,但自由性和蔓延性有明顯下降,表明城市在快速發(fā)展的同時逐漸向緊湊型、集約型方向靠攏。從區(qū)域擴張的總體情況來看,研究區(qū)有較高的總體自由度和蔓延度,但兩者都呈減緩趨勢,此外其擴張優(yōu)勢度較低,表明區(qū)域土地仍未得到有效利用。 利用研究區(qū)DEM數(shù)據(jù)與不透水面信息進行疊加分析。從高程分布看,絕大部分的不透水面位于1803米到2022米高程范圍內(nèi),其擴張趨勢為由發(fā)展趨于飽和狀態(tài)的低海拔區(qū)域逐漸轉(zhuǎn)向開發(fā)難度較大的較高海拔地區(qū),但目前低海拔區(qū)域仍是主要擴張區(qū)域。從坡度分布看,絕大部分的不透水面位于2°到15°范圍內(nèi),并表現(xiàn)出不透水面面積隨坡度的上升而下降的總體趨勢。2-6°為主要擴張區(qū)域,15°以上的土地建設(shè)開發(fā)難度明顯增加,范圍內(nèi)不透水面面積增長較少。 選取斑塊類型和景觀水平指數(shù)進行研究區(qū)景觀格局分析。從景觀水平看,研究區(qū)斑塊總數(shù)由2002年的63395上升到2006年的77244繼而下降到2010年的69629,區(qū)域經(jīng)歷了斑塊破碎加劇進而逐漸聚合的過程。其中,水體、植被和夾雜稀疏植被的裸地部分的斑塊數(shù)為先上升后下降,無植被覆蓋的裸地部分的斑塊數(shù)呈持續(xù)下降趨勢。區(qū)域內(nèi)不透水面斑塊數(shù)先下降后上升,其經(jīng)歷了從眾多離散分布的較小面積到逐漸趨于集中的較大面積斑塊區(qū)域的變化過程,此外其在發(fā)展過程中與其他景觀的臨接關(guān)系減弱,毗鄰的景觀類型趨于單一,較大程度影響了區(qū)域生態(tài)多樣性。
[Abstract]:The city is the most since twentieth Century the characteristics of human activities significantly, with the rapid growth of population and the rapid development of science and technology, human transformation speed and the demand for land of city urbanization increasingly grow with each passing day,. The impervious surface as the characteristic product of the high-speed city in the process of human transformation of surface activity, rapidly increasing its area of great influence on the regional ecological environment evolution process and dominate the city landscape pattern. As the matrix landscape in the city and the typical surface types, impervious surface is often used for land cover dynamic monitoring, city hydrological simulation, estimation of non-point source pollution load, thermal environment effect, regional climate change, many aspects of city expansion monitoring and the city ecosystem evaluation.
This paper takes Kunming city as the study area, remote sensing, GIS and landscape pattern analysis software support, the improved decision tree method to extract impervious surface information and using the Pearson chi square test was used to analyze the spatial expansion of quantitative Shannon entropy statistical analysis model, from the aspects of policy guidance, terrain analysis and driving force of expansion the analysis of regional landscape pattern from the city ecological landscape level. The study area in 20022006 and 2010 three Landsat ETM+ images based on the image, select the 3,4,5,7 band and the image enhancement processing principal component analysis in the process of the first band and normalized difference vegetation index after computing the vegetation information enhancement band common 6 band image using decision combination. The tree classification method based on CRUISE algorithm of land cover classification interpretation, and by using the improved normalization index and based on The impermeable surface density partition map of vegetation coverage is used for mask processing to further improve the accuracy of interpretation, and the classification information of land use in the three period of the study area is obtained.
The study area is divided into 8 regions using direction space quadrant method, using Pearson chi square test, Shannon entropy, dominance, expansion intensity index analysis model from three aspects of the development direction of regional, stage and the overall situation is studied through analysis of surface area quantitative spatial expansion. The results show that the direction of the impervious the water expansion rate and expansion trend differences, reflects the imbalance of.2002 ~ 2006 years of city spatial expansion, impervious surface expansion by the policy, planning and other factors are less constrained, but the expansion rate of free development is relatively slow.2006 ~ 2010 years, its growth rate has accelerated the expansion degree has obvious advantages the promotion, but free and spreading has decreased significantly, that in the rapid development of the city gradually to the compact, close to the intensive direction. From the overall situation of regional expansion, research The region has a higher overall degree of freedom and spread, but both show a slow trend, in addition to its low expansion advantage, it shows that the regional land has not been effectively used.
Using the DEM data of the study area and impervious surface information overlay analysis. From the view of the distribution of elevation, most of the impervious surface is located 1803 meters to 2022 meters elevation range, its expansion trend from low altitude regional development tends to saturation gradually turned to the difficulty in the development of higher altitude, but the low altitude region is still the main expansion area. From the view of the distribution of slope, the vast majority of impervious surface is located at 2 degrees to 15 degrees range, and showed the general trend of the impervious surface area decreased with the slope of the rise.2-6 degrees as the main expansion area, more than 15 degrees of land development is a significant increase, impervious surface area increased less range.
Patch type and landscape level index of landscape pattern analysis. From the landscape level, the total number of patches in the study area increased from 63395 in 2002 to 77244 in 2006, then dropped to 69629 in 2010, the region has experienced the process of fragmentation intensified and gradually polymerization. In the water, vegetation and bare soil part with sparse vegetation the number of patches was increased after the first drop, no part of the bare patch number of vegetation coverage showed a declining trend. Within the area of impervious patch number decreased and then increased, the experience from many discrete distribution is relatively small to the change process of large area plaque area gradually tends to focus on the development of the process and the other landscape adjacent relationship weakened, adjacent landscape types tend to be single, greatly influenced the ecological diversity of the region.

【學位授予單位】：昆明理工大學
【學位級別】：碩士
【學位授予年份】：2013
【分類號】：P237

【參考文獻】

相關(guān)期刊論文 前10條

1 吳萍;姜黎霞;;熵的泛化及應用[J];北京聯(lián)合大學學報(自然科學版);2005年04期

2 章文波,符素華,劉寶元;目估法測量植被覆蓋度的精度分析[J];北京師范大學學報(自然科學版);2001年03期

3 劉珍環(huán);王仰麟;彭建;;不透水表面遙感監(jiān)測及其應用研究進展[J];地理科學進展;2010年09期

4 孫志英;趙彥鋒;陳杰;李桂林;檀滿枝;;面向?qū)ο蠓诸愒诔鞘械乇聿豢赏杆忍崛≈械膽肹J];地理科學;2007年06期

5 楊勝天,劉昌明,楊志峰,孫睿,周曉東;南水北調(diào)西線調(diào)水工程區(qū)的自然生態(tài)環(huán)境評價[J];地理學報;2002年01期

6 匡文慧;劉紀遠;陸燈盛;;京津唐城市群不透水地表增長格局以及水環(huán)境效應[J];地理學報;2011年11期

7 周紀;陳云浩;張錦水;李京;;北京城市不透水層覆蓋度遙感估算[J];國土資源遙感;2007年03期

8 姜紅梅,任立良,安如,袁飛,王美榮;基于土地利用與地表覆蓋遙感信息的洪水過程模擬[J];河海大學學報(自然科學版);2004年02期

9 徐涵秋;;城市不透水面與相關(guān)城市生態(tài)要素關(guān)系的定量分析[J];生態(tài)學報;2009年05期

10 李德仁,王樹良,李德毅,王新洲;論空間數(shù)據(jù)挖掘和知識發(fā)現(xiàn)的理論與方法[J];武漢大學學報(信息科學版);2002年03期

，

本文編號：1444350