Providing a Clearer Picture of Arctic Change with Drone and Piloted Airborne Remote Sensing Data

The Science

Widespread changes in vegetation cover and composition in the Arctic tundra impact Arctic ecosystems, people, and climate feedbacks. While tundra landscapes are highly heterogeneous, quantifying the distribution and composition of tundra vegetation over large extents has been challenging. Pixels from traditional satellite observations are too coarse and contain a mixture of plant species that are hard to differentiate. To address this, researchers combined high-resolution unoccupied aerial system (UAS) data and airborne imaging spectroscopy to estimate pixel-wise composition of Arctic plant functional types (PFTs) with NASA’s Airborne Visible/Infrared Imaging Spectrometer–Next Generation (AVIRIS-NG).

The Impact

UAS data were capable of characterizing species composition in heterogeneous tundra landscapes with overall mapping accuracy greater than 86%. The UAS-derived vegetation maps were also highly effective as training data for upscaling and mapping vegetation composition across watersheds with AVIRIS-NG imagery. This approach produced accurate composition maps (5 m) of 12 PFTs at spatial extents large enough for effectively monitoring tundra vegetation changes and informing modeling efforts focused on improving Earth system predictability. Collectively, this study demonstrates the utility of UAS platforms for providing training data for developing cutting-edge multiscale approaches needed to fill gaps in understanding of Arctic regions.

Summary

Widespread changes in vegetation cover and composition in response to ongoing climate changes in high latitudes are responsible for significant effects on Arctic ecosystem functioning and global climate feedbacks. However, accurately quantifying the composition and distribution of tundra vegetation over large areas is challenging given that commonly used satellite observations are too coarse to differentiate low-lying tundra vegetation types.

To address this challenge, researchers combined airborne observations from very high-resolution (VHR, ~5 cm) UAS platforms and hyperspectral imagery from NASA’s AVIRIS-NG instrument to develop novel multiscale methods to map the fractional composition of 12 key low-Arctic tundra PFTs. Using high-resolution vegetation maps developed with novel UAS imagery as training observations, new statistical partial least squares regression (PLSR) models were developed to predict the continuous fractional cover of each PFT with AVIRIS-NG imagery. The PSLR models’ performance was evaluated using additional UAS data reserved from model training and against other traditional methods used to map vegetation fractional coverage within remote sensing pixels.

These methods showed that: (1) a wide range of Arctic PFTs can be mapped using VHR UAS imagery with an overall accuracy greater than 86%; (2) generated UAS maps can then be used effectively as training data for larger-scale models developed with airborne AVIRIS-NG imagery with a mean absolute error <0.13; and (3) the final AVIRIS-NG PLSR models outperformed traditional linear mixture analysis. These new scaling approaches could likely be transferred to other Arctic regions where similar data is available. These approaches could also potentially transform understanding of fine-scale patterns in tundra vegetation composition, improve long-term monitoring of tundra vegetation dynamics, and improve process-based modeling of Arctic tundra ecosystems.

Principal Investigator

Shawn Serbin
Brookhaven National Laboratory
[email protected]

Co-Principal Investigator

Daryl Yang
Brookhaven National Laboratory
[email protected]

Program Manager

Daniel Stover
U.S. Department of Energy, Biological and Environmental Research (SC-33)
Environmental System Science
[email protected]

Funding

This work was supported by the Next-Generation Ecosystem Experiments-Arctic (NGEE-Arctic) project that is supported by the Biological and Environmental Research (BER) program in the U.S. Department of Energy’s (DOE) Office of Science, and through DOE contract No. DE-SC0012704 to Brookhaven National Laboratory. Support was also received from NASA’s Future Investigators in NASA Earth and Space Science and Technology (FINESST) Grant 80NSSC22K1296 and the NASA Surface Biology and Geology Mission Study (#80GSFC22TA016).

References