| Based on theory and methodology of landscape ecology, the landscape configuration of the 3 cities along the Changjiang Ricer, Maanshan, Wuhu and Anqing in Anhui province, were analyzed by using the technique of RS, GPS and GIS. The paper is the first report on comparison of landscape pattern in different cities in the same region in this country, which will be referenced for Chinese urban forest constructing. The paper mainly concluded three aspects.1) Based on the remote sensing data collected from 2002TM image and 2003 Quickbird satellite photograph of the three cities, the author created a Geography Information System of different urban landscape elements using GIS, and then the data was determined using GPS in field survey to make an adjustment.The characteristic indices for landscape pattern, such as landscape diversity, landscape dominant, landscape evenness, patch shape index, the density of landscape patch, the density of landscape element patch, landscape isolation index and recreation forest accessible index, were applied to describe the landscape pattern of urban forest in three cities.The results showed that there were 5 major landscape elements (excluded road), i.e. building and hard pavement surface, urban forest, farmland, waters and general green belts in three cities. The landscape matrix was building and pavement surface. The percentage urban forest patch occupied in the 3 cities based on land area was as the following in order: Maanshan 16.74%, Anqing 14.41%, and Wuhu 10.43%. According to land use issue and its function, the urban forest could be classified as the following 6 types: landscape forest, yard woods, recreation woods, road side trees, ecological and shelter forest. According to the patch area, the urban forest could be classified as the following 5 types, small patch (<500m2), middle patch (501 m2~2000 m2), large patch (2001 m2~10000 m2), big patch (10001 m2~50000 m2) and huge patch (>50000 m2). Based on this class, the landscape diversity of urban forest was, Maanshan 1.58, Anqing 1.74, and Wuhu 1.56; The area proportion of urban forest patches with more than 1hm2 reached to 78.9% of whole urban forest area in Maanshan, then 82.8% in Anqing, and 77.8% in Wuhu respectively, which showed that the large urban forest patch occupied dominant status in three cities; Landscape isolation index of urban forest is 1.21 in Maanshan, 1.13 in Anqing and 1.42 in Wuhu ; and the density of landscape element patch of urban forest is 0.89 in Maanshan, 0.72 in Anqing and 0.81 in Wuhu, which indicated that fragmented trait of urban forest was high, connected trait was low, and the distribution of urban forest was not balance. In order to appraise the recreation function of urban forest, the author created recreation forest accessible index, which would indicate if recreation woods were convenience for city dwellers to use; The result showed that Maanshan reached to 49.7%, Anqing 42.3%, and Wuhu 36.7%.2) To detect the gradient of urban forest pattern, author conducted a series of analyses along a transect cutting across the entire city area from suburb to downtown on satellite photograph according the direction of urbanization. In the paper, the gradient variations were described with the landscape indexes of continuous samples along the transect. The author attempted to catch the relation between human disturbances and urban forest landscape pattern. The results showed that urbanization intensity could be reflected by the changes in urban forest landscape pattern along transect from suburb to the center of downtown The gradient variation of urban forest landscape pattern was complex following a curve with many apexes, and had its own spatial trait, in all three cities. Based on the gradient analysis, The urban forest could be classified as the 4 categories along transect: urban forest in commercial area of urban center, urban forest in industry and resident square, urban forest in link region between suburb and building up area, urban forest in suburb. The urban forest area had a higher proportion of land in city zone, it follows that the level of urban forest constructing is good in the center of downtown for these three cities. Because of the differences of geography location, urban properties and land use pattern as well, each city had its owe trait of urban forest landscape pattern along transect.To study the variation degree of urban forest landscape pattern along transect, the author put forward other two new landscape indexes, the distribution variation index in suburb-urban area (URS)and the distribution variation index (US)in city zone. The result of URS was 44.19 in Maanshan, 118.62 in Anqing and 18.29 in Wuhu. The result of US was, 18.61 in Maanshan, 31.71 in Anqing and 17.33 in Wuhu.3) Combined the subjective and objective method, the author set up the comprehensive assessment system on urban forest landscape structure, which was employed to make a practically assessment for the tree cities. The system included three levels: aim level (A) - urban forest landscape structure; rule level (B): urban forest landscape compositions (B1), urban forest landscape pattern; index level(C): urban forest coverage (C1), urban forest area for per capita (C2), urban forest landscape diversity (C3), urban forest landscape isolation index (C4), the density of urban forest landscape element patch(C5), recreation forest accessible index (C6), urban forest distribution variation index in city zone (C7). Based on the system, the three cities were assessed with three different methods to decide the importance of different level indexes. The three ways all showed the same sequence of the index of urban forest landscape structure: Maanshan>Anqing>Wuhu. The result indicated that the Maanshan had the best urban forest landscape pattern, then was Anqing and Wuhu successively. the best Compared to the three ways for deciding the importance of different level indexes, the author thought that the method-"Expertise Consultation +CRITIC"was the best.
|
PDF Full Text Request