| Wetlands are almost all over the world, but the real research of wetlands science is a matter for nearly half a century. Though the current definition of wetlands is not exactly the same at home and abroad, the《onvention on Wetlandshas》been the definition of the various national and international research institutions to accept, "wetlands means no matter natural or artificial, permanent or temporary, moor,swampland,former wetlands, peat bog, or a static or mobile, or for fresh water, brackish water or salt water of the body of water, including water depth at low tide does not exceed 6m in shallow water regions." Wetlands'types includs marine/coastal wetlands, inland wetlands, constructed wetlands three main types, which include permanent shallow waters of coastal rocks, sand, beach, lakes, herbaceous peat, alpine wetlands, tundra wetlands, shrub wetlands, aquaculture ponds, irrigated land, salt, canals, drains, and the Coastal Zone region of coral reefs, sea grass areas, such as mangroves and estuaries. As a result of both terrestrial ecosystems and aquatic ecosystems of the characteristics of wetlands are the most productive on earth special ecological types, with maintaining the water sources, water purification, flood and drought control and regulate the climate, landscape and maintenance of biological diversity and other important ecological functions. Because of these functions, Wetlands are known as the "kidney of the Earth," "natural reservoir" and "natural species banks"This paper focused on the Yancheng coastal wetland area of information extraction techniques of remote sensing of wetland resources, as well as changes in landscape analysis and prediction. We research a simulated reflectivity image processing method, the amount of information integrated in each band, and has reduced losses in the information and the application of science and energy characteristics of radiation to simulate the reflectance map to achieve, so that the shadow area of image quality, improved resolution to improve the overall image quality and clarity to distinguish features, include the seven-band TM data information, access to information integrity. Remote sensing images in the process of visual perception physiology, which based wetland monitoring technology, establish a simulated reflectivity of TM multi-band image the best information, the optimal band combination of monitoring TM453. Remote sensing images which based on color, texture, shape features, the growth of knowledge, GIS attributes, spatial relations and a series of stimulating knowledge and learning aids such as the fact that the use of information and integrate into a variety of earth science knowledge, increased knowledge acquisition, knowledge maintenance and knowledge reasoning, such as matching operation and the achievement of a range of issues. The use of the study area frome 1988 to 2009 in a total of five time wetland information to extract data, combined with the study model of compositive intelligent model to get the main land changes in the distribution of cover types, the information extraction accuracy and Kappa coefficient of the whole better than the BP neural network classification of the overall accuracy of 84.33%, Kappa coefficient of 0.821 and maximum likelihood supervised classification method for the classification of the overall accuracy of 81.21%, Kappa coefficient was 0.804, results as follows:in April 1988 the total accuracy of 89.12%, Kappa coefficient was 0.889; in June 1994 the total accuracy of 89.97%, Kappa coefficient was 0.875; in June 2000 the total accuracy of 90.11%, Kappa coefficient was 0.897; September 2006 the total accuracy of 89.41%, Kappa coefficient of 0.871; in March 2009 the total accuracy of 90.27%, Kappa coefficient was 0.888. With the idea of landscape ecology, spatial pattern of the use of index of Jiangsu Province, Yancheng wetland resources of the dynamic changes of landscape pattern analysis studies. We study Markov model for the purpose of study area predict the pattern of development trends, the forecast results show that in the next 10 years, the type of non-wetland landscape of the area will become increasingly large proportion of residents and aqua farm will remain an increasing trend, while the cropland area will be a decreasing trend after 2006, the development of urban sprawl, major transportation, power, energy, communications, etc. take up much of the implementation of the project area of arable land. Over time, natural wetland resources Seepweed, reed, Spartina patens only exist in two nature reserve.Yancheng wetland area will be the progressive development of constructed wetlands aqua farm are becoming the major wetlands and other wetland resources gradually reduce regional trends.
|
PDF Full Text Request