Development of integrated wetland change detection approach: In case of Erdos Larus Relictus National Nature Reserve, China.

摘要: Abstract Wetlands are threatened by the global warming and human exploitation pressure, have been shrinking quickly in recent years. Timely accurate wetland area change detection is primary task for conservation restoration. The objective of this study to develop an integrated approach which integrates advantages spectral mixture analysis (SMA) vector (CVA) identification dynamics. In proposed approach, water, vegetation soil fractions wetlands were derived SMA; then, detailed information (including magnitude 12 direction categories) calculated through CVA. was applied Erdos Larus Relictus National Nature Reserve (ELRNNR), China, using time-series Landsat images during 1977–2017. We found that faced serious degradation, with water fraction changed (5.79 km2), (1.35 km2) both (3.53 km2). From 1977 2000, a slight degradation occurred northeast edge Bojiang Lake marginal Houjia Lakes inside ELRNNR, vegetation. During 2000–2010, severe from 2010 2017, more susceptible precipitation activities. Analysis result indicated long-term drought effects mismanagement as well misuse beings driving factors degradation. achieves higher accuracy than classification detect change, ability obtain information.