Dynamics And Prediction Of Landscape In Land Utilization Of Nanjing City

In this dissertation Nanjing City was chosen as the study case a and its land information system during1989-2009was built on the remote sensing image in1989,2000and2009and the statistical data was obtained via GIS technology. The status and changes of land use types were analyzed and the landscape characteritics furtherly was summarized by the calculation of landscape indicators. The trend of land use types during2011-2017was conducted by Markov chain model finally. The main conclusions were:(1) Cultivation land is the dominant landscape in Nanjing City. During the study period, the area of cultivation land, woodland, grassland, aquatorium (water area) and others decreased. The grassland area decreased in a rapid rate while the construction land area increased on the contrary. Land alteration occurred between each paired land use types, among which the transition from cultivation land to woodland and the inter-transition between construction land and water area were much more frequent. Compared with the period of1989to2000, the period of2000to2009experienced a more intensified transition.During the study period the number of patches (NP), patch density (PD) and edge density (ED) of urban landscape in Nanjing City increased while the mean patch size (MPS) decreased, indicating the acceralated disturbance from human beings. At the same time the fragmentation of patches increased and the connectivity between patches decreased. Under the rapid urbanization, the area-weighted mean shape index(AWMSI) gradually diminished, along with the complexity of patch form. The area-weighted mean patch fractal dimension (AWMPFD) slightly increased while Shannon’s diversity index (SHDI) and Shannon’s evenness index (SHEI) both increased, which illustrated a more evenly distribution of all types of patches and the prominence of landscape heterogeneity.(2) For cultivation land, woodland, water area, construction land and others, NP kept increasing while for grassland NP increased in the first stage (1989-2000), followed by a reduction in the second stage (2000-2009). Construction land had the highest NP among all land use types. During the study period NP of cultivation land had the maximum variation, followed by woodland in the second. Since PD and NP correlated positively, construction land had the highest PD. followed by cultivation land and woodland. Cultivation land had the highest total edge (TE) and ED. The decreasing of the cultivation land area and the increasing of its TE and ED showed that more other land used types scattered across cultivation land, thus enhancing the extent of fragmentation. The increasing speed of TE and ED for woodland during2000-2009was not as fast as that during1989-2000, indicating the distribution of woodland became more concentrated. TE and ED for grassland increased during1989-2000and then decreased at a faster pace during the second period, indicating a quicker transition of grassland to other land use types. Along with NP results and land transition matrix it could be conclued that the increment of TE and ED for construction land were chiefly caused by the growth of the number of patches during1989-2000and by both the growth of patch number and the enlargement of patch area during2000-2009. Water area had both a higher ED and a larger TE along with less area and more patches, demonstrating the transition of water area to other landscape as well as land fragmentation. The area for others was relatively small so the variation of TE and ED during the study period was not obvious.Fractal dimension (FRACT) value for cultivation land was larger due to the complexity of patch form than that for construction land, woodland, grassland, water area and others. The variation of FRACT value for construction land was slight could be explained by the urban plan oriented layout. The value of MSI and mean patch fractal dimension (MPED) showed the same variation trend.(3) Preliminary stage,landscape fragmentation of Jiangning and Central district were more than Liuhe、Pukou、Gaochun and Lishui. The results of AWMSI and AWMPFD indicated:Jiangning and Liuhe had the most complication shape and least man-created landscape; the rest of subareas had the less complication shape which Central district had the least shape and the most man-made interference. SHEI and SHDI illustrated:land types of Central district and Jiangning was more abundance; By contrast, land types of Liuhe and Pukou were single; Lishui and Gaochun were in place in the middle.Last stage, patches entire of Central district and Pukou were more than Gaochun、Lishui、Jiangning、 Liuhe. The results of AWMSI and AWMPFD indicated:Liuhe had the most complication shape, Lishui、Jiangning、Pukou、Gaochun and Central district were subsequently. SHEI and SHDI illustrated:land types of Jiangning and Lishui was more abundance; By contrast, land types of Central district、Pukou、Liuhe and Gaochun.(4) The period of2000to2009, NP increased inordinately in all sub areas, which Jiangning、Liuhe、Lishui and Gaochun were more increasing and the less increasing among Central district and Pukou because of a large parts construction land. MPS and NP degrees of change were opposition. In the research period, AWMSI of Liuhe、Central district and Lishui increased than Jiangning and Gaochun. In the same time, SHEI of Central district and Gaochun decreased slightly,other sun areas were increasing. Expansion of Nanjing City built-up area was lack of uniformity in the research period. During1989-2009, because policy-mandated and geographical factors limited, built-up area expanded slowly. The21st century,built-up area was rapid expansion, average expansion area and expansion strength significantly increased. Between twenty years, expansion in Nanjing City was highest standard in history.In the process of built up areas expansion, fractal dimension increased and space forms increasingly complicated.(5) The prediction of landscape during2011to2017revealed the increasing in the area of construction land and water area at the price of reduction of the other four landscapes. Although the Markov chain model had room for improvement the forecasting results still could be helpful for the comprehensive understanding of landscape variation trend. Despite the great effort the government had paid in cultivation land protection, land consolidation, land reclaimation and planning, the area of cultivation land would still be decreasing in the near future.