Long-term Remote Sensing Dynamic Monitoring And Ecological Assessment Of Coastal Wetland In Hangzhou Bay

Coastal wetlands are complex ecosystem with enormous ecological and economic value.Due to urbanization,excessive reclamation,environmental pollution and other issues,the coastal wetlands have undergone dramatic changes in recent decades.They are facing serious risk of degradation,impaired ecological service functions,affecting coastal ecological balance and sustainable economic development.Long-term sequence dynamic monitoring of coastal wetlands and assessing changes in ecological values and security,which helps to understand the long-term evolution of coastal wetland ecosystems and their driving mechanisms,and provide scientific basis for the rational protection of coastal wetland resources and formulation ecological restoration policies.Hangzhou Bay coastal wetland were selected as the study area,a new proposed method were estiblished to map coastal wetland distribution based on high-resolution remote sensing images,and dynamic change of coastal wetland were monitored by time series Landsat remote sensing data.Then,the ecological value of the coastal wetland in Hangzhou Bay was assessed based on the principles of ecology and economics.The improved TOPSIS model were cereated by entropy method,and was used to evaluate the ecological security status of the coastal wetland in Hangzhou Bay,and the grey prediction model was used to predict and analyze the ecological security in 2020.The main research contents,results and conclusions are following:(1)A new proposed approach for mapping coastal wetland land covers based on the object-oriented and expert knowledge.After analyzing the spatial distribution and spectral characteristics of different wetland types in the study area,the decision tree method,based on the expert knowledge,phenologyical characteristics,was use to identify and extract land cover types in Hangzhou Bay.The overall classification accuracy of new proposed methods is 90.28%,and the kappa coefficient is 0.89.Meanwhile,this new method based expert knowledge have better performance than the maximum likelihood and Random forest method.The results shows that the new proposed method combined with phenological characteristics can effectively enlarge the separability of different land cover types,and it is suitable for accurate classification in complex coastal wetlands.(2)Based on time series Landsat data,the dynamic change of coastal wetland land cover was monitored in Hangzhou Bay,and the temporal expand process of coastal wetland was analyzed.The results showed that,the land expanded to the sea year by year;the coastal wetlands(Scirpus triqueter salt marsh,Spartina alterniflora salt marsh,freshwater herbaceous marsh and aquaculture pond)expanded outward and disappeared continuously on the inland side(converted to other types);The area of impervious surface increased rapidly,mainly to the north.The analysis of typical sites shows that the evolution and driving factors are different along the coast.Wetlands evolved under the combination of natural succession and human activities.The main driving forces are natural accumulation,artificial reclamation and urban construction.The main factors are terrain,hydrology,soil,population,economy,policy,etc.(3)Based on the method of ecology and economics,combined with the results of remote sensing land cover,the ecological value of Hangzhou Bay coastal wetland were evaluated in 2000,2005,2010 and 2015.The total value of ecosystem services for Hangzhou Bay coastal wetland in2000,2005,2010 and 2015 was 413.01×10~8,261.25×10~8,259.52×10~8 and 224.3×10~8 RMB,showing a decreasing trend.The main service category of Hangzhou Bay coastal wetland is the regulating service,the value of which accounted for more than 60%of the total ecosystem service value.The main functions of Hangzhou Bay coastal wetland are climate regulation,water purification and soil conservation,the values of which account for more than 80%of the total ecosystem service value.Over-reclamation and urban construction reduced wetland area,resulting in the ecosystem values decrease.(4)The ecological security status of the coastal wetlands in Hangzhou Bay was evaluated and predicted,based on the Driving force-Pressure-State-Impact-Response(DPSIR)conceptual model.The ecological safety values were calculated in 2000,2005,2010 and 2015,and the value of 2020was predicted.The result shows that ecological security values of Hangzhou Bay in 2000,2005,2010 and 2015 were 0.413,0.382,0.287 and 0.582,respectively.The security level deteriorated from the warning level to the vulnerable level,and returned to the warning level,showing an upward trend after the decline.The grey prediction model predicted that the ecological security value of the Hangzhou Bay coastal wetland in 2020 is 0.697,which is in a“relatively safe”state.(5)The driving mechanism of long-term evolution of coastal wetlands is qualitatively analyzed,which under the combined effect of natural succession and human activities.It is found that hydrological conditions,topography and species invasion are the main natural driving factors,and population,economy,policy,reclamation,urban construction and aquaculture are the main human driving factors.From the perspective of coastal wetland ecology,this paper forms a set of technical methods of extracting complex land cover in coastal wetland ecosystem by expert knewledge&phonological charcacteristic,monitoring dynamic changes of coastal wetland by time series remote sensing data,constructing ecosystem service system of coastal wetland,evaluating and predicting ecological environment and ecological security.The results propose new ideas and methods for remote sensing ecological dynamic monitoring and evaluation of coastal wetland ecosystems,providing theoretical basis for scientific and rational selection of ecological restoration measures,and have important significance for rational utilization of coastal wetland resources in Hangzhou Bay and sustainable socio-economic development.