Data accompanying: "Snow persistence lowers and delays peak NDVI, the vegetation index that underpins Arctic greening analyses"

This dataset contains multispectral orthomosaics generated from repeat drone imagery captured during a single growing season at one Arctic and one sub-Arctic tundra site. The orthomosaics are used to calculate the persistence of fine scale snow cover in:

Hoad et al. (in press) - Snow persistence lowers and delays peak NDVI, the vegetation index that underpins Arctic greening analyses

All code to analyse and visualise this data can be found in the following GitHub repository: https://github.com/calumhoad/snowpersistence

Data summary:

This dataset contains an orthomosaic for each date where a drone survey was conducted at three tundra plots: Blæsedalen (BL), Kluane Low (KL) and Kluane High (KH). The Blæsedalen data were captured using the integrated sensor aboard a DJI Mavic 3 Multispectral and the Kluane data were captured using a SAL Engineering MAIA-S2 sensor. For more information on methdology, please refer to our Environmental Research Letters article and supplementary materials. The imagery for each plot is listed below:

• Blæsedalen (BL):
  • 2023-07-02
  • 2023-07-12
  • 2023-07-18
  • 2023-07-26
• Kluane Low (KL):
  • 2022-06-29
  • 2022-07-05
  • 2022-07-18
  • 2022-08-01
  • 2022-08-14
• Kluane High (KH):
  • 2022-07-09
  • 2022-07-19
  • 2022-07-29
  • 2022-08-04
  • 2022-08-13

To replicate analyses from the manuscript:

Clone the GitHub repository before downloading this data, then place the contents of each subfolder from the dataset into the following folders of the GitHub repo:

• For Blæsedalen files: data/uav/orthomosaics/m3m/5cm
• For Kluane Low files: data/uav/orthomosaics/maia/kluane-low/5cm
• For Kluane High files: data/uav/orthomosaics/maia/kluane-high/5cm

Cite as:

Hoad, C., Myers-Smith, I.H., Kerby, J.T., Colesie, C. and Assmann, J.J., (in press). Snow persistence lowers and delays peak NDVI, the vegetation index that underpins Arctic greening analyses.

Abstract (from manuscript):

Satellite imagery is critical for understanding land-surface change in the rapidly warming Arctic. Since the 1980s, studies have found positive trends in the normalised difference vegetation index (NDVI) derived from satellite imagery over the Arctic—commonly referred to as ‘Arctic greening’ and assumed to represent increased vegetation productivity. However, greening analyses use satellite imagery with pixel sizes ranging from tens to hundreds of metres and do not account for the integration of abiotic phenomena such as snow within vegetation indices. Here, we use high-resolution drone data from one Arctic and one sub-Arctic site to show that fine-scale snow persistence within satellite pixels is associated with both reduced magnitude and delayed timing of annual peak NDVI, the base metric of Arctic greening analyses. We found higher snow persistence within Sentinel-2 pixels is associated with a lower magnitude and later peak NDVI, with a mean difference in NDVI of 0.1 and seven days between high and low snow persistence pixels. These effects were stronger in NASA HLSS30 data, representative of Landsat data commonly used in greening analyses. Our findings indicate that unaccounted changes in fine-scale snow persistence may contribute to Arctic spectral greening and browning trends through either biotic responses of vegetation to snow cover or abiotic integration of snow within the estimated peak NDVI. In order to improve our understanding of Arctic land-surface change, studies should integrate very-high-resolution data to estimate the dynamics of late-season snow within coarser satellite pixels.

Acknowledgements (from manuscript):

We would like to thank everyone who helped with field data collection in the Canadian Yukon during 2022 and in Greenland during 2023, including Joseph Everest, Erica Zaja, Jiri Subrt, Sian Williams and Mariana García Criado. For help with drones and sensors, particular thanks go to Tom Wade at the University of Edinburgh Airborne Research and Innovation facility, and Alex Merrington, Jack Gillespie, Craig Atkins and Robbie Ramsay at the NERC Field Spectroscopy Facility. Additional thanks to Alan Hobbs, Colin Kay and Graham Mitchell from the NERC Geophysical Equipment Facility.

We thank Tim Gyger for support and consultation on our statistical methods, Gwenn Flowers for the time taken to provide climate data for Kluane and Kirsten Schmidt-Pedersen for sharing her extensive knowledge of the people, plants and animals of Qeqertarsuaq, Greenland.

Funding for this research was provided by NERC through a SENSE CDT studentship (NE/T00939X/1), the NERC Tundra Time project (NE/W006448/1), a 2023 UK-Greenland Arctic Bursary, a NERC Geophysical Equipment Facility loan (1152), and a NERC Field Spectroscopy Facility loan (891.0111). Additional funding was provided by a Scottish Alliance for GeoScience, Environment and Society (SAGES) small grant scheme award to Calum Hoad.

We thank Kluane First Nation and Champagne and Aishihik First Nations for their permission to work on their lands. We gratefully acknowledge the people of Kalaallit Nunaat in general, and of Qeqertarsuaq in particular, for being able to conduct this research on their land. We thank Outpost Research Station and Arctic Station for logistical support.

The authors acknowledge constructive comments from two anonymous reviewers, which greatly improved the manuscript.

Author contributions (from manuscript):

Calum G. Hoad: Conceptualisation (lead); Data curation (lead); Formal analysis (equal); Funding acquisition (equal); Investigation (lead); Methodology (lead); Project administration (lead); Resources (lead); Software (equal); Visualisation (lead); Writing – original draft (lead); Writing – review and editing (lead). Isla H. Myers-Smith: Conceptualisation (supporting); Formal analysis (supporting); Funding acquisition (equal); Methodology (supporting); Resources (supporting); Supervision (lead); Visualisation (supporting); Writing – review and editing (equal). Jeff T. Kerby: Conceptualisation (supporting); Methodology (supporting); Supervision (equal); Visualisation (supporting); Writing – review and editing (equal). Claudia Colesie: Conceptualisation (supporting);  Funding acquisition (supporting); Supervision (equal); Project administration (supporting); Resources (supporting); Visualisation (supporting); Writing – review and editing (equal). Jakob J. Assmann: Conceptualisation (supporting); Data curation (supporting); Formal analysis (equal); Investigation (supporting); Methodology (supporting); Supervision (lead); Software (equal); Visualisation (supporting); Writing – original draft (supporting); Writing – review and editing (equal).