Bridging Arctic pathways: integrating hydrology, geomorphology and remote sensing in the North

Abstract

This work presents improved approaches for integrating patterns and processes within hydrology, geomorphology, ecology and permafrost on Arctic landscapes. Emphasis was placed on addressing fundamental interdisciplinary questions using robust, repeatable methods. Water tracks were examined in the foothills of the Brooks Range to ascertain their role within the range of features that transport water in Arctic regions. Classes of water tracks were developed using multiple factor analysis based on their geomorphic, soil and vegetation characteristics. These classes were validated to verify that they were repeatable. Water tracks represented a broad spectrum of patterns and processes primarily driven by surficial geology. This research demonstrated a new approach to better understanding regional hydrological patterns. The locations of the water track classes were mapped using a combination method where intermediate processing of spectral classifications, texture and topography were fed into random forests to identify the water track classes. Overall, the water track classes were best visualized where they were the most discrete from the background landscape in terms of both shape and content. Issues with overlapping and imbalances between water track classes were the biggest challenges. Resolving the spatial locations of different water tracks represents a significant step forward for understanding periglacial landscape dynamics. Leaf area index (LAI) calculations using the gap-method were optimized using normalized difference vegetation index (NDVI) as input for both WorldView-2 and Landsat-7 imagery. The study design used groups to separate the effects of surficial drainage networks and the relative magnitude of change in NDVI over time. LAI values were higher for the WorldView-2 data and for each sensor and group combination the distribution of LAI values was unique. This study indicated that there are tradeoffs between increased spatial resolution and the ability to differentiate landscape features versus the increase in variability when using NDVI for LAI calculations. The application of geophysical methods for permafrost characterization in Arctic road design and engineering was explored for a range of conditions including gravel river bars, burned tussock tundra and ice-wedge polygons. Interpretations were based on a combination of Directcurrent resistivity - electrical resistivity tomography (DCR-ERT), cryostratigraphic information via boreholes and geospatial (aerial photographs & digital elevation models) data. The resistivity data indicated the presence/absence of permafrost; location and depth of massive ground ice; and in some conditions changes in ice content. The placement of the boreholes strongly influenced how geophysical data can be interpreted for permafrost conditions and should be carefully considered during data collection strategies.