| Summary: | Under the pressures of amplified Arctic warming, developing detailed spatiotemporal understandings of Arctic surface water systems is an urgent task. This thesis establishes a method for monitoring inundation in lakes and wetlands in the Mackenzie River Delta, using Sentinel-1 SAR. The importance of dual-polarimetric entropy and speckle filtering were evaluated in a set of random forest models. Results indicated that integrating entropy and omitting the speckle filter during SAR pre-processing may be promising for surface water modelling in the Mackenzie River Delta. Analysis of lake-level inundation variability revealed spatial clusters of variable and stable lakes and suggested that lake variability is sensitive to discharge volume of the Mackenzie River. The lake-tracking framework presented in this thesis can act as an important tool for evaluating the long-term hydrological response of Arctic ecosystems, to a rapidly changing climate. Ontario Graduate Scholarship Natural Sciences and Engineering Research Council of Canada
|
|---|