GEOSCAN Search Results: Fastlink

GEOSCAN Menu


TitleIntegration of multi-scale remote sensing data for reindeer lichen fractional cover mapping in eastern Canada
 
AuthorHe, L; Chen, WORCID logo; Leblanc, S GORCID logo; Lovitt, JORCID logo; Arsenault, A; Schmelzer, I; Fraser, R HORCID logo; Latifovic, R; Sun, L; Prévost, C; White, H PORCID logo; Pouliot, D
SourceRemote Sensing of Environment vol. 267, 112731, 2021 p. 1-12, https://doi.org/10.1016/j.rse.2021.112731
Image
Year2021
Alt SeriesNatural Resources Canada, Contribution Series 20200631
PublisherElsevier
Documentserial
Lang.English
Mediapaper; digital; on-line
File formatpdf; html
ProvinceNewfoundland and Labrador; Quebec
NTS12M; 13D; 13E; 22I; 22J; 22K; 22N; 22O; 22P; 23A; 23B; 23C; 23G; 23H
AreaLabrador; Réservoir Manicouagan; Smallwood Reservoir; Labrador City
Lat/Long WENS -70.0000 -62.0000 54.0000 50.0000
Subjectsgeophysics; Nature and Environment; Science and Technology; remote sensing; satellite imagery; photogrammetric surveys; mapping techniques; ecosystems; environmental studies; climate effects; modelling; reindeer lichen; Cladonia spp.; WorldView; Landsat; Methodology; Wildlife; drones; Climate change; machine learning; Artificial intelligence
Illustrationslocation maps; satellite images; digital images; flow charts; plots; models
ProgramCanada Centre for Mapping and Earth Observation
Released2021 10 07
AbstractReindeer lichens (Cladonia spp.) are an essential food source for caribou especially during winter. They can also be a valuable indicator for ecosystem health and climate change. Inventory of lichen abundance at regional scales is required to assess availability within caribou ranges, and assess potential declines from natural and anthropogenic disturbances. Previous studies have mapped lichen cover and volume using remote sensing, but these efforts were often constrained by the limited availability of ground truth information needed for model calibration and validation. In this study, we leveraged unoccupied aerial vehicle (UAV) surveys and WorldView (WV) satellite scenes in a nested upscaling approach in order to expand the number of training samples at the 30 m Landsat resolution. These were used to develop machine learning models to map fractional reindeer lichen cover in Eastern Canada. We found that the best correlation between UAV and WV derived lichen coverages exists at an optimal scale that is slightly larger than 30 m and varies with landscape type and observation geometry. Based on training data from UAV-calibrated lichen coverage from WV data, a neural network model with simple structure achieved a root mean square error (RMSE) = 0.09, a mean absolute error (MAE) = 0.07 and R2 = 0.79 for mapping fractional lichen cover from Landsat without the use of ancillary data. We then applied our model and Landsat data to produce a lichen fractional cover map for the Red Wine Mountain caribou herd range in Labrador, NL and the Manicouagan caribou herd range in Québec. Validation against domain-averaged lichen cover in eight UAV survey sites suggests an accuracy with RMSE = 0.04, MAE = 0.03 and R2 = 0.62 for low lichen cover. Compared to aggregated lichen cover at 30 m from UAV surveys, map accuracy decreases to RMSE = 0.09, MAE = 0.06, and R2 = 0.49, partially due to registration error between UAV and Landsat images. Our study demonstrates that upscaling of lichen cover from UAV data to Landsat via an intermediate image scale is an effective regional-scale mapping approach.
Summary(Plain Language Summary, not published)
Reindeer lichens are an essential food source for caribou especially during winter. They can also be a valuable indicator for ecosystem health and climate, and global changes. Recently, concerns arise as their abundance in caribou habitats declines due to natural and anthropogenic disturbances. Inventory of lichen conditions at regional scales and its accessibility to caribous is required to address these issues. In this study, we developed a model to map lichen cover by combining drone aerial surveys, very high-resolution satellite, the Landsat data at 30 m resolution and machine learning method. Our method achieved a relatively high accuracy for mapping fractional lichen cover from Landsat without the use of additional data. We then applied our model to produce a lichen fractional cover map for the Red Wine Mountain caribou herd range in Labrador and the Manicouagan caribou herd range in Québec. Our study suggests that combining remote sensing data at different spatial scale is an effective mapping approach. Our lichen cover map will be used for assessment of caribou forge quality and to understand the caribou mitigation behaviors.
GEOSCAN ID327899

 
Date modified: