Study On Coastal Wetland Ecosystem Health Warning Theory

The coastal wetland is a coastal area,which is located at the juncture of the ocean and the land,and has the characteristics of fresh water and salt water.At the same time,it is also the transition zone of the two ecological systems of land and sea.Coastal wetland natural environment,there are rich species,high productivity,and contains a large number of valuable mineral resources,such as oil and so on.Resistance in the marine disasters,improve the local microclimate,beautify the surrounding natural geographical environment,seasonal runoff and flood regulation,maintaining the entire coastal wetland area has a very important role in the ecological balance.The coastal wetland ecosystem is not the health of the main attention is that the entire wetland system for a certain space and time scale of the pressure to be able to carry out elastic recovery,and to maintain stability,and no major functional diseases in the system,the system is complete in a variety of key organizations.With the development of society,economy,science and technology,with the population growth of coastal residents rapidly.The coastal wetland ecosystem safety has become hot issue everywhere,the ecological system of wetland area is the growing threat to health,if things go on like this may threaten,the development of human beings,therefore,for the establishment of wetland ecological system health warning has become a priority among priorities,to find out the causes from the change of wetland ecosystem health,establish early warning system of ecological health,through the data that the future health,thus more scientific way of management and provide scientific basis for the level of the wetland ecosystem health and recovery sustainable management.This paper first defines the basic theory of coastal wetland ecosystem health warning required,in rough set theory(RS)and BP(BP ANN)neural network theory and method,from the Liaodong Bay has selected five representative areas: Jinzhou City,the city of Ling Hai,Dawa County,Panjin city and Panshan County,based on the "driving force-pressure-state-response-control" of this framework to select the early warning indicators,first use 18 indexes to construct the index model,and then through the reduction function of rough set theory reduction out duplicate index,finally screened out 4 indicators,leaving the index system of 14 evaluation indexes the construction of the early warning model;their weight and clear one grade and two grade index by AHP method,then finally standardized by comprehensive index method to judge the wetland system early warning situation.The results showed that: the health status of the wetland ecosystem in the five study areas was not very good.From 2000 to 2009,during the nine years,warning situation of each region have increased a level,Linghai City,Jinzhou,Panjin City is a heavy district.Finally,the BP neural network model was used to predict the wetland ecosystem health in five areas in the next nine years.The results show that the results are not optimistic.The resulting conclusions and need all strengthening of Liaodong Bay coastal wetland natural environment protection and governance,a step by step to promote the recovery and growth of the coastal wetland ecological,reach a win-win situation for improving the economy of people in Liaodong Bay coastal wetland ecological protection and coastal areas.