Hakai Institute’s Kelp Drone Mapping Plan

Download a fillable version of the Kelp Drone Mapping Template:

Background

The Hakai Nearshore kelp monitoring program focuses on tracking in situ changes in diversity and productivity of giant kelp (Macrocystis pyrifera) and bull kelp (Nereocystis luetkeana) kelp habitats and their associated communities; determining the drivers of change in BC kelp systems; and understanding connectivity between kelp and other nearshore habitats. The monitoring program started in 2014 to track seasonal variability of understory and giant and bull kelp forests. A refined set of scuba surveys now takes place annually in a subset of sites around the Calvert region on the Central Coast of British Columbia, Canada.

The Hakai Marine Habitat Mapping Program aims to document interannual variation in kelp forests, seagrass meadows and rocky intertidal habitats in British Columbia at local, regional and coast-wide scales using a variety of remote sensing tools appropriate to different spatial scales. The first local-scale drone surveys for kelp forest monitoring were conducted in 2015 and have since been expanded to several other sites along the Central Coast (see Figure 1). Research and development is focused on the application of drone mapping of kelp beds for mapping kelp extent and biomass at the species level. Along with pioneering machine learning work to automate kelp forest mapping processes, work is currently underway to link surface and subsurface kelp metrics and examine predictive abilities of modelling kelp biomass.

Together, the Hakai Nearshore kelp monitoring program and Marine Habitat Mapping Program work together to study kelp forest ecosystems in British Columbia using both in situ and spatial mapping techniques. 

Hakai Institute Kelp Monitoring Program  

1. What objectives do you wish to accomplish through kelp monitoring?

The Hakai Institute’s goals for kelp drone mapping are threefold: (1) track changes in extent and distribution and productivity of Macrocystis pyrifera and Nereocystis luetkeana kelp habitats; (2) determine the drivers of change in BC kelp forest ecosystems; and (3) mobilize datasets and methodologies to support coastal science and practitioners. This work supports long-term ecological research to develop maps of kelp presence, link surface and subsurface kelp metrics, and map kelp extent and biomass at the species level across the Central Coast. 

Drone mapping provides high-resolution aerial imagery that captures kelp canopy extent and characteristics at a spatial scale appropriate for site-level monitoring (10-100s of meters). This technology enables us to: (1) Document seasonal and inter-annual changes in kelp bed extent; (2) Distinguish between canopy-forming kelp species (Macrocystis vs. Nereocystis) using RGB and multispectral imagery; (3) Link surface canopy metrics with subsurface productivity data collected through in situ dive surveys; and (4) Ground-truth satellite imagery for regional and coastwide kelp forest mapping. By pairing drone surveys with subtidal monitoring, we can identify correlations between environmental drivers and kelp forest dynamics, ultimately improving our ability to predict and understand kelp distribution patterns across the BC’s Central Coast.

2. Based on your goals, which metrics are appropriate for you to monitor?

With drone we capture the following metrics: (1) Kelp extent – the spatial distribution and total area of kelp canopy cover at each survey site; (2) Kelp species – distinguishing between Macrocystis pyrifera (giant kelp) and Nereocystis luetkeana (bull kelp) using high resolution drone imagery; and (3) Kelp biomass – estimates derived from linking drone-based canopy measurements with in situ biomass data. These metrics are essential for tracking changes in kelp forest distribution and productivity, and understanding drivers of change.

3. How do you plan to measure and monitor those metrics?

The Hakai Institute conducts annual drone surveys at several long term monitoring sites on the Central Coast of British Columbia (Figure 1). Sites were selected to be representative of bull kelp (Nereocystis luetkeana) and giant kelp (Macrocystis pyrifera) conditions on the Central Coast, appropriate for long term ecological research. Site selection criteria included: (1) Presence of established kelp beds with known Macrocystis or Nereocystis dominance; (2) Accessibility for both drone operations and paired in situ dive surveys; and (3) Representativeness of different habitat conditions across the Calvert region (as is safe for surveys).

Drone surveys are conducted at low-tide (<1.5 m) during peak summer biomass to capture maximum annual extent of kelp forests. All surveys are done flown from a stationary vessel as shore access at monitoring sites is limited and/or dangerous. Paired diver surveys happen within 1-3 weeks of drone surveys and capture subtidal measurements of biomass, frond density, plant counts, and productivity data as well as oceanographic data.  

Survey Design and Field Logistics

1. What resources (people, time and equipment) do you need? 

Team members

Time requirements 

1. Field work: Boat-based drone kelp surveys happen over 1-2 weeks at 13 sites on the Central Coast in July of each year. Monitoring sites range in size (0.035 km2 - 0.2 km2) and each one is paired with a diver monitoring site. Sites are sized to capture representative kelp bed dynamics while remaining within practical drone flight time and data processing constraints. Note: kelp drone surveys are paired with seagrass monitoring surveys which happen at the same time.

2. Data processing: Depending on site size and complexity, data storage, processing and completion of analysis-ready data can take anywhere from 3-5 hours per site. The development and release of the Habitat Mapper, a machine learning tool which automates kelp canopy detection from drone imagery in 2021, greatly reduced processing times for analysis ready data. 

3. Analysis and Report: Time for analysis and reporting varies annually but several days are allocated for analysis and internal and external reporting (e.g. summary reports, Annual Research Reports for First Nation partners). 

Equipment requirements

• Drone (e.g. Mavic 3E) 

• Micro SD cards

• Table/phone controller (if not incorporated into drone)

• GPS + GCPs if required

• VHF receiver

• Camera, field notebook, pen
  

Funding for this program is through the Tula Foundation. 

Supporting metrics

What additional or supporting metrics, if any, do you need to collect to meet your project’s objectives?

Our program integrates multiple data sources to provide comprehensive kelp monitoring: (1) high resolution satellite imagery and fixed-imagery for regional- and coast-wide scale kelp mapping and to ground-truth larger-scale patterns; (2) In situ dive surveys that collect understory kelp biomass, urchin populations, and detailed productivity metrics; and (3) oceanographic data collected in situ and derived from remote sensing data sources.

Physical oceanographic parameters are collected through collaboration with Hakai’s Nearshore and Oceanography programs. Measurements include temperature, chlorophyll, nutrients, and dissolved inorganic carbon (DIC) to identify environmental drivers of kelp growth and distribution.

Species presence and ecosystem diversity are assessed through paired in situ dive surveys conducted as part of the Hakai Nearshore program. These surveys document fish and invertebrate diversity, understory kelp species, grazers (urchins, snails), and ecosystem indicators including seastar wasting disease.

Human impacts are considered within the broader context of our long-term monitoring research. While not explicitly tracked as separate metrics in our drone mapping program, our research network includes collaborations that address human factors affecting kelp systems, including: sea otter recovery and predation dynamics, potential harvest impacts (both historical urchin harvest and kelp harvest), and broader marine management considerations. Our site selection in the Calvert region provides baseline data for a relatively low-impact area, which can be compared with other regions experiencing different human pressures.

Data Management Plan

Our data management follows FAIR (Findable, Accessible, Interoperable, Reusable) principles to ensure long-term value and accessibility of kelp monitoring data. Key objectives include: (1) Maintaining high-quality, well-documented datasets that support long-term ecological research and predictive modeling; (2) Making data discoverable and accessible to researchers, managers, and Indigenous communities through the Hakai Data Catalog and through direct communication and sharing; (3) Ensuring data interoperability with other kelp and nearshore monitoring programs; (4) Supporting reproducible science through transparent data processing and analysis workflows; and (5) Preserving data for future analyses and comparison with historical baselines.

Data are archived in standardized, open formats to ensure long-term accessibility and interoperability. Raw drone imagery is stored as jpegs, orthomosaics as GeoTIFF files with full spatial referencing. Processed data products including spatial datasets of kelp extent with species classifications provided in GeoPackage and shapefile formats for GIS compatibility. Tabular data (site characteristics, survey metadata, quantitative metrics) are archived in CSV format. All data packages released through the Hakai Data Catalog follow metadata standards and include comprehensive documentation to support data reuse.

Data security is managed through Hakai Institute’s data management infrastructure with redundant backup systems and version control. Kelp location and extent data are generally non-sensitive and suitable for public release, though specific site locations may be considered in consultation with local First Nations and managers if there are concerns about harvest pressure or other impacts. Data are shared openly through the Hakai Data Catalog with appropriate licensing (Creative Commons Attribution 4.0) to enable broad use while ensuring attribution. Specific data sharing agreements may be established for collaborative research projects with academic or management partners.

Roles and responsibilities concerning data

Data roles and responsibilities are clearly defined: The Hakai Geospatial Team is responsible for data collection (drone flights), processing (imagery analysis, kelp classification), and initial quality control. Luba Reshitnyk (Habitat Mapping Program Lead) oversees data management planning and coordinates with Ondine Pontier (Nearshore Program Lead) to integrate drone data with in situ monitoring data. The Hakai Data Services team manages data archiving, metadata creation, and publication through the Hakai Data Catalog. Data are openly available to researchers, managers, and the public through the catalog, with usage tracked for reporting and program evaluation.

Long term storage of data

Long-term data storage and preservation are managed through the Hakai Institute’s data infrastructure with multiple redundant backup systems. Processed data products are published through the Hakai Data Catalog (https://catalogue.hakai.org/dataset/ca-cioos_291b98a4-d868-462c-852a-d6cf79ecf6ce) with DOI assignment for permanent citation and discovery. Raw and intermediate data products are maintained in Hakai’s internal data management systems with regular backups. The Hakai Data Catalog is part of the Canadian Integrated Ocean Observing System (CIOOS) network, ensuring discoverability through national and international marine data portals. Data preservation follows best practices for long-term archival including regular integrity checks, format migration as needed, and comprehensive metadata to support future use.