Seasonal Inundation Dynamics on the Barrow Peninsula, AK

Abstract:

It is important to know how surface evapotranspiration and change in inundated areas are correlated, especially in flat Arctic wetlands such as the tundra region near Barrow, Alaska, as their underlying frozen ground and low hydrological gradient due to flat relief confine the lateral runoff of their standing water. Moreover, knowledge regarding seasonal dynamics of inundated areas is expected to be an essential and controlling factor in modeling regional energy and hydrological balance, which are related closely to frozen ground stability, in Arctic wetlands. However, the seasonal change and spatial distribution of inundated areas have not yet been well explored and quantified.

Here we’ve deployed high spatial resolution (WorldView2 and QuickBird) images of Barrow area on eight dates from 2006-2014, to investigate seasonal change of inundated areas for a 4700 ha wetland, including the Barrow Ecosystem Observatory. Inundation dynamics were measured in the field in 2014 using DGPS. These ground truth data was used to develop a classification algorithm for discriminating between open water, overgrown water (mixed vegetation and standing water), and dry surfaces in the high-resolution images. The inundation index is created by combining NIR band, NDVI, and stack mean of BGR and NIR bands, and shown to be capable for mapping the extent of open water, dry, and overgrown water surfaces. In order to explore the relationship between water balance and changes in the inundated area, the estimated seasonal change in the inundated areas was compared with the daily surface water balance (rainfall – evaporation) calculated using available micrometeorological data for the years 2006-2014.

Our results suggest that inundation dynamics correlated with the surface water balance during mid-late summer (July-September), though this relationship was not valid in the early summer (June), when surface hydrology is governed mainly by surface runoff above the shallow thawing front of the ground. With the inundation index developed and relationship between inundation index and surface water balance quantified in this study, it will become possible to automatically estimate inundation dynamics, which will improve our understanding of Arctic wetlands hydrology and support the scaling of local measurements to regional scales.