Spectral Characteristics of Vegetation Functional Traits across a Range of Thaw Gradients on Alaska’s Seward Peninsula
Wednesday, 17 December 2014
The Arctic and Boreal regions are warming rapidly, leading to the thawing of the underlying permafrost and associated changes in vegetation structure and composition. The thawing of ice-rich permafrost drives land surface dynamics called thermokarst, characterized by a variety of geomorphic surface features across high latitude landscapes. The development of these thermokarst or thermo-erosional features depends on factors such as local permafrost conditions, hydrology, geomorphology, vegetation, and climate, but their degree of dependence are not well understood across scales. The structure, functions and traits of the vegetation can work as effective indicators of these landscape changes. Our ability to characterize these vegetation characteristics across a wide range of thaw gradients at the local scale could help us to better understand the dependency as well as the impacts of thermokarst processes on them. This will also help us to develop capabilities to quantify these characteristics and dependencies from local to regional scales by using remote sensing and ecosystem modeling techniques. During the months of June - July of 2013 and 2014, we conducted field surveys at various sites across the central Seward Peninsula in Alaska covering a range of thaw gradients to collect data for vegetation functional traits, ancillary data and also hyperspectral data in the 400-2500 nm range using a field spectrometer. Data were collected from plots established along 50 m transects to capture transitional states of these thaw features from the upland zone, transition zone, and thaw lake basins as well as in polygonal features. Here we discuss the characteristics of vegetation functional traits and how they relate to the ground-based spectral measurements. Some of these findings could be scaled up using airborne and satellite remote sensing data. The findings from this study can improve our understanding of disturbance patterns and their feedbacks to local scale plant and soil dynamics. Scaling up our understanding based on multi-scale remote sensing and ecosystem models over multiple spatial and temporal scales across landscapes could help us reduce uncertainties in estimating the carbon budget from local to pan-arctic scales.