Extent of Canopy-Forming Kelps, Derived from World View-2, Central Coast, Central Coast, British Columbia

This dataset represents the extent of canopy-forming kelp (Nereocystis luetkeana and Macrocytis pyrifera) present at the ocean surface (hereafter referred to as surface canopy-forming kelp) for the year and season the imagery was collected. Canopy-forming kelp (Order Laminariales) are found along the coast of British Columbia. Two species, bull kelp (Nereocystis luetkeana) and giant kelp (Macrocystis pyrifera) are the most common species and are ecologically and biologically important components of nearshore ecosystems. Mapping their distribution is important for understanding the state and health of coastal areas.

The WorldView-2 satellite was tasked by the Hakai Institute and imagery collected in the summer of 2014 for the Goose group, McMullin group, Calvert Island, and Stirling Island region. Additional WorldView-2 imagery was acquired via a DigitalGlobe Foundation Grant for the South Calvert region, Simonds Group, Admiral and Tribal Groups region, Spider Island and Cultus Sound region. Imagery from the Stryker Island and Athlone Island region was provided through a collaboration of the Department of Fisheries and Oceans Canada and the Hakai Institute.

Prior to classification, each image region was corrected for atmospheric effects and land areas were masked. A supervised classification (maximum liklihood algorithm) was conducted to classify pixels which represent surface canopy-forming kelp. The results were imported into ArcGIS (10.5.1) and manually edited to remove pixel regions that were misclassified as kelp.

Attributes: Region: A description of the geographic region of each image from which surface kelp extent was derived. Shape_Area: extent of surface kelp in meters squared. Tide_m: Approximate tide height at time of image acquisition in meters above chart datum

Methodology: The extent of surface canopy-forming kelp was derived from WorldView-2 imagery. WorldView-2 is a multispectral sensor with 8 spectral bands which are coastal (400-450 nm), blue (450-510 nm), green (510-580 nm), yellow (585-635 nm), red (630-690 nm), red edge (705-745 nm), near infrared 1 (770-895 nm) and near infrared 2 (860-1040 nm). WorldView-2 imagery has a pixel size of 2 × 2 m. Previous research has demonstrated that, like terrestrial vegetation, kelp reflects light in the near infrared portion of the electromagnetic spectrum which makes it detectable by multispectral imagery (Cavanaugh et al. 2010).

For each image scene, an atmospheric correction was applied using the ATCOR algorithm within Geomatica (PCI Geomatica, 2016). Settings for ATCOR were: aerosol type = maritime, condition = mid-latitude summer, constant visibility = 100 km. Land regions were removed from the image scene in ArcGIS (10.5.1). An additional 5 m buffer from the masked land was used to remove erroneous signals from pixels close to shore (e.g. shallow submerged rocks, shadows, waves, etc). A supervised classification (maximum likelihood algorithm) was conducted in ENVI 5.3 (Harris Geospatial). A minimum of 100 pixels were sampled for each class. Classes include deep water, shallow water, glint, and surface kelp. A shadow class was also used in cases where shadows from onshore features such as trees were still present in the scene. After classification, a median filter (3x3) and a sieve filter (which removed classified pixels less than 3 pixels in size) were used to reduce noise and smooth the dataset. The classified kelp pixels were exported as a shapefile. Additional editing of misclassified kelp was conducted within ArcGIS 10.5.1.

Cavanaugh K, Siegel D, Kinlan B, Reed D. Scaling giant kelp field measurements to regional scales using satellite observations. Mar Ecol Prog Ser. 2010;403:13-27. doi:10.3354/meps08467.

Technical details: Vector data of surface canopy kelp extent derived from WorldView-2 imagery.

The majority of the imagery was collected August 13th, 2014. The image scene centered around the Stryker Island and Athlone Island region was collected July 22nd, 2012. The image scene centered around the South Calvert region was collected June 4th, 2015.

This dataset was derived from WorldView-2 multispectral imagery. Atmospheric correction of images was conducted in Geomatica (PCI Geomatics, 2017). Land masking was done in ArcGIS (10.5.1). Supervised classification for determining kelp surface extent was conducted using ENVI 5.3 (Harris Geospatial). Classification outputs were edited within ArcGIS 10.5.1.

Contributors: Luba Reshitnyk and Wiebe Nijland

Cite as: Canopy-forming kelps extent derived from WorldView-2 imagery, Central Coast, British Columbia, Canada. Version 1.0. Hakai Institute. Dataset. [access date].)

Access and Use

Licence: Appropriate credit must be given to Hakai Institute and the authors of the dataset.

Data and Resources


Metadata Created October 26, 2018, 22:48 (UTC)
Metadata Updated October 26, 2018, 22:48 (UTC)
Reference Date(s) 2017-12-23 (Creation)
2017-12-23 (Publication)
Frequency of Update
Metadata Date March 14, 2018, 21:08 (UTC)

Graphic Preview

Dataset extent

Map data © OpenStreetMap contributors

Additional Info

Field Value
Contact Email data@hakai.org
encoding utf8
metadata-language eng
progress completed
resource-type dataset
Responsible Party Hakai Institute (Point of Contact, Processor)
spatial-reference-system 26909