| Wetland,which is the special natural integration of wet creatures,wet environment,forests and oceans form the three global ecosystems.It is an important natural resource and is one of the most important environments for human survival as well.It not only offers the raw materials for human life and human production but also plays an important role in improving weather condition,flood storage,pollution control and reducing other environmental pollution.As the study area at the junction of the three countries,poor accessibility and other conditions can be limited,the site investigation is difficult.Traditional research method of wetland cannot meet the requirements.So,remote sensing technology becomes the most ideal method of investigating wetlands and monitoring their dynamic change.In this study,we took the Tumen River estuary trans-boundary areas as the study areas.Using the stratified classification method of incorporating geoscience knowledge,we extracted wetlands information of the Tumen River estuary trans-boundary areas in 1980-2015.Then,we carried out wetland condition monitoring and wetland change analysis about Tumen River estuary trans-boundary areas,and conducted a comparative study.The conclusions of the research are got1?A new rule or judge function is brought forward to improve the image classification result,in view of the stratified classification method.Provided by the non-remote sensing information,such as NDWI,named as a normalized Difference Water Index,and NDVI,named as a Normalized Difference Vegetation Index,are respectively used to formulate rules for the image classification.And vegetation non-vegetation classification was carried out using SVM,neural network and maximum likelihood classification.We use the decision tree classification method to carry out the waters.The results showed that the stratified classification method of incorporating geoscience knowledge is suitable for Tumen River estuary trans-boundary areas wetland classification.The classification method is generally applicable to the study of Landsat series image wetland classification in the same period.2?In the analysis of the dynamic changes of the Tumen River estuary trans-boundary areas,the wetland area showed a decreasing trend,although The wtland area was slightly mitigated from 1988 to 1999?The reduction of natural wetlands was more severe than that of artificial wetland.And,the wetland space changed significantly in China Jingxin Town,DPR Korea Ende County,Seonbong County and Russia Hassan Town.The temporal and spatial changes of the North Korea's wetlands had a greater impact on the overall changes in the Tumen River estuary.The reduction of marsh wetland area was the main form of wetland change in this period.The increase in paddy fields and reservoirs also resulted in a decrease in natural wetlands and an increase in artificial wetlands.In Tumen River estuary of China,the main phenomenon was that the ponds changed to the lakes,lakes changed to reservoir.And swamp,paddy fields changed to woodland,dry land.In Tumen River estuary of DPR Korea,The phenomenon of conversion to dry land and Bare land was more serious.In Tumen River estuary of Russia,nature wetlands,such as lakes,rivers,swamps,changed to meadow.Meadow was the main flow of wetland transformation in Tumen River estuary of Russia.There were obvious differences in landscape pattern in Tumen River estuary trans-boundary areas.In the degree of fragmentation,the wetland was the weakest in Russia,followed by wetlands in DPR Korea and the largest in Tumen River estuary of China.In the degree of dominance,the dominant landscape of wetland in China is not obvious,the swamps dominance was declining and paddyfields was rising.Lakes landscape was the dominant landscape in DPR Korea,but affected by swamps and paddy fields change.Lakes and swamps landscape was the dominant landscape in Russia.Uneven distribution of the area is the main reason why landscape consistency of china was higher than that of the other two countries.3?The wetland change of Tumen River estuary trans-boundary areas was the result of comprehensive influence of natural driving factors and human driving factors.From a long time scale,we found that affected by the regional oceanic climate and the characteristics of the terrain,the wetlands are mostly distributed in the lower coastal terrain,mainly in the territory of Russia,China Jingxin Town and around the surrounding areas of DPR Korea's Lakes.And from a short time scale,limited by DPR Korea's and Russian's humanities data acquisition difficulties and other conditions,we found that Human drivers are undoubtedly the main drivers of wetland change.By the different national policy system,the differences of economic development,the differences of population growth rates and the construction of water conservancy projects and other human factors,wetlands show different changes between China,DPR Korea and Russia.Finally,some research work that needs to be carried out in the future is pointed out as regards to the perfection of the image classification model and the wetland change analysis model. |
PDF Full Text Request