Abstract
Permafrost peatlands are responding to recent high-latitude climate warming in dramatic fashion. These changes in terrain surface characteristics are affecting hydrology in a variety of ways. Increasing summer precipitation is leading to top-down thaw of permafrost across a variety of ecotypes. At smaller scales, studies are reporting the expansion of lateral thaw features and increased rates of thermokarst formation. Surface water plays a critical role in these processes. We have been combining site level field measurements, geophysics, remote sensing, and machine learning geospatial analyses to establish connections between the snowpack, vegetation, and permafrost thaw. The relationships we have identified allow projection of our site scale measurements across broader regions. This presentation summarizes results of recent studies by our research group at a variety of Interior Alaska peatland sites. In the first study, of the seasonal snowpack, we combined airborne hyperspectral and LiDAR measurements with machine learning methods to characterize relationships between ecotype and more than 26,000 snow end of winter snowpack measurements. We focused from 2014-2019 at three field sites representing common boreal ecoregion land cover types. These winters represent anomalously low (2016), typical mean, and high (2018) snowpacks. Hyperspectral measurements account for two thirds or more of the variance in the relationship between ecotype and snow depth. An ensemble analysis of model outputs using hyperspectral and LiDAR measurements yielded the strongest relationships between ecotype and snow depth. Since the seasonal snowpack often provides more than half of the yearly water equivalent these results have ramifications for surface water dynamics. In another study we used Landsat products to estimate fire-induced thaw settlement across the ice-rich Tanana Flats lowland in Interior Alaska that contains fens, bogs, and a variety of other wetland features. After linking fire areal extent, burn severity, land cover changes, and post-fire vegetation recovery we developed an object-based machine learning ensemble approach to estimate fire-induced thaw settlement from comparing repeat LiDAR to Landsat products. Our model delineated thaw settlement patterns across six unique fire scars and explained ~65% of the variance in LiDAR-detected elevation change. Results from a long term study of fen hydrology and climatology across Tanana Flats has tracked changes to hydrologic features and thermokarst development using historical image analysis, site scale measurements, and ground based geophysics. Repeat electrical resistivity tomography and high resolution ground surface elevation measurements identified thaw subsidence at a 10 year fire scar of more than a meter as a result of up to three meters of top-down permafrost thaw. At the same sites we have been able to quantify how lateral thaw of permafrost has led to the expansion of small ponds and bogs. We are now working to combine these geophysical and survey measurements with remote sensing information to project these land cover changes over a larger spatial extent.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have