| Human activities as the dominant factor in the regional landscape pattern change occupies very important position in the global environment change. With the acceleration of urbanization, the landscape pattern had become one of the hot spots in the current research. The paper taked Nanchang city as the research area, with the theory of landscape ecology as the guide, with the "3S" technology as the means, using the three phase Landsat series image map and the forest resources survey data, vegetation coverage was estimated based on pixel two model with TAVI, and the terrain gradient variation of landscape vegetation coverage, and response to the change of landscape pattern were studied. From four dimensions that contain landscape structure change, the transfer matrix, the landscape pattern index and landscape change information map, the spatial and temporal evolution characteristics of landscape pattern were analyzed, and the topographic gradient effect of landscape change information mapwas also analyzed. Based on the Spatial statistical method, the optimal analysis scale was determined, and the space-time variation characteristics of landscape patterns were analyzed, and the effect of the change of landscape pattern and landscape change behavior was analyzed. The SD system was constructed, and the non-space drive was simulated based on the causal relationship between factors of landscape pattern change. Then the GIS-logistic coupling model was constructed to analyze the space drive of the landscape pattern. Finally, the ecological control measures were put forward. The results for the study area provide scientific basis for ecological restoration measures, meanwhile, it enriched the theory and method of landscape pattern. Therefore, the thesis research has important theoretical and application value. The main research results and conclusions as follows:(1) The vegetation cover of the study area in 2010, 2005 and 2001 were 0.54, 0.45 and 0.42 respectively. High vegetation coverage area total area ratio decreased from 76.82% to 53.24%, middle vegetation coverage area steady at 19%, middle-low and low coverage area increased from 13.17% to 21.20%, and from 7.96% to 18.99%. The forest vegetation cover in 2001, 2005 and 2010 were 0.72, 0.62 and 0.58 respectively, high coverage area total area ratio decreased from 72.25% to 40.07%, T middle vegetation coverage area increased from 13.72% to 16.65%, middle-low and low coverage area increased from7.02% to15.05%, and from 5.18% to 19.60%. The results show that the vegetation coverage and forest vegetation coverage in Nanchang city in the 2001-2010 years showing the characteristics of the evolution of high coverage to middle-low and low cover. The vegetation cover was degraded, but the degradation rate decreased after 2005. The correlation coefficient of vegetation cover and height in the three stage was 0.94, 0.97 and 0.95 respectively, and the correlation coefficient with the slope was 0.97, 0.98 and 0.98 respectively. The positive correlation between vegetation cover and height, slope gradient was obvious, and gradient variation characteristics was obvious. The change of landscape pattern had a great influence on the change of vegetation coverage, and landscape utilization in the vegetation coverage change area had been changed nearly 40%.(2) From the view of landscape structure, landscape types in the study area were cultivated land and forest mainly, forest land is dominated by coniferous forest. In the 2001~2010 years, the cultivated land and grassland decreased continuously, and the construction land and forest land increased continuously, and the other land and water area had a negative fluctuation. From the view of landscape type transfer matrix, the Landscape outflow was dominated by farmland, water, other land and forest land, and the landscape inflow was dominated by farmland, other land, construction land and forest land. Landscape patches from 5584 to 7773, and diversity index increased from 1.31 to 1.41, and dominance degree reduced from 0.27 to 0.21, and the fractal dimension increased from 1.33 to 1.36 in 2001-2010. From the change of landscape pattern index, the landscape had a distinct tendency of fragmentation. From the perspective of landscape change information map analysis, the main landscape pattern was "stable type", accounting for 56.47% of the total area. Among them, "cultivated land-cultivated land- cultivated land " was the important TUPU map. Landscape change information map showed obvious gradient effect. Continuous changes, transitional changes, early and late stage changes varied with the change of terrain gradient. The distribution in the 4-5 level terrainwas dominant position, and the distribution index decreased in the 6-8 level. The stable and the other four types showed a reverse change, which were the dominant distribution area besides 4-6 level.(3) The variability of the landscape pattern vulnerability under the 2.5km×2.5km sampling scale was significant, exponential model was fitting model. The ratio of Structure variance to the total variance was 0.71, 0.55 and 0.53 respectively. There was a strong spatial correlation. The evolution of landscape pattern in 2001-2005 and 2005-2010 in two periods showed the opposite characteristics. Respectively, the evolution of the middle-low and middle vulnerability to middle-high and high vulnerability, and the evolution of high and middle-high vulnerability to the middle and middle-low vulnerability. The I Moran’s index of landscape vulnerability was 0.7687, 0.8516 and 0.258 respectively. It showed that the spatial distribution of landscape pattern had a strong positive correlation, and was a clustering distribution pattern. The absolute coefficient(R2) of regression model for the change of vegetation cover change and landscape pattern were 0.7639 and 0.7265, and confidence reached 99.99%. The results showed that the change of landscape pattern is highly correlated with the change of vegetation cover degree, and the change of landscape pattern can reflect the changes of regional ecosystem.(4)The analysis of the response of landscape pattern evolution and ecosystem showed that the stable pattern had a great contribution to the reduction of the landscape pattern and the stability with contribution rates were 58.45% and 56.19%. The contribution to the increase of the vulnerability of the stage changes and the transition changes were more with 15.50% and 13.57%. The impact of the contribution to the change of vulnerability was smaller while the area of continuous changes was smaller. contribution of "cultivated land-cultivated land- cultivated land to degrees of vulnerability was the largest among stable type. Among of typical landscape change mode, " deforest and reclamation " and " reclain lake bottom land and plant in to crops" had higher contribution rate to the increase of vulnerability, and "returning farmland to forest", "abandoned land", "land development", "returning farmland to Lake","building occupied" and "land reclamation" had contribution to lreduced contribution. The results showed that the evolution of landscape pattern has important influence on the changes of ecosystem, and the reasonable utilization and protection of landscape had important significance to maintain the stability of landscape ecosystem.(5) From the point of view of the bearing capacity of each landscape type, the system dynamics model was builded, considering the total supply of grain, economic development and ecological environment construction, the demand for farmland, forest land and construction land were considered. The SD model reflected the causal relationship and positive and negative feedback of the driving factors accurately. Through the statistical survey data and simulation test method to determine the function of each factor. The relative error between the simulation value and the actual value was within 5%. The results showed that the SD model could simulate accurately non spatial driving factors on the landscape pattern evolution, aslo could simulate the demand of landscape types Under the different conditions.(6) Spatial models of landscape type transformation were constructed based on the GIS-logistic coupling model with better fitting degree. The explanatory power of the driving factors on the landscape pattern change in space was analyzed nicely. The results showed that the main explanatory variables which changed to farmland included the land development, annual rainfall, GDP, soil nutrients, and distance from the residential areas. The main explanatory variables which changed to forest included the forest land cover, distance from the river, slope, "returning farmland to forest" and slope. The main explanatory variables which changed to construction land included the distance from the residents, the fixed assets investment, Elevation, the total population and the distance from the road. The results showed that the social and economic factors and human disturbance caused by the man-made effects had a short and significant impact on landscape pattern. |
PDF Full Text Request