Essay

Beluga whales have nutritional, economic, social, and cultural importance for Inuit in Inuit Nunangat. Beluga whales are harvested for subsistence in communities across the Arctic. To ensure harvesting is sustainable, regular population abundance and distribution assessments of beluga whale populations is essential for ensuring stable populations for years to come. Traditional abundance estimates of marine mammals have been conducted by boat-based or fixed-wing aircraft aerial surveys. These surveys are logistically challenging in remote areas, they require careful planning, a large team of people, and have the potential to disturb marine mammals. Aerial surveys can also pose risks to field personnel associated with flying in the Arctic including landings in remote locations, changes in elevation, and potential rapid changes in weather (e.g., fog). The evolution of remote sensing methods has promoted its use as a non-invasive and safer alternative to aircraft surveys for assessing animal abundance and distribution.

The beluga whale image shown here was taken by satellites and drones in the Churchill River on July 31, 2022. Here, beluga whales are shown swimming, seemingly undisturbed. It showcases the relatively non-invasive potential for monitoring marine mammals using drones and satellites. Throughout this essay I delve into using remote sensing methods for the monitoring of marine mammals from a science-based lens. Remote sensing methods have expanded our ability to regularly monitor the surface of the Earth. The wealth of available imagery has evolved significantly since the first camera was invented in 1816 by Joseph Nicéphore Niépce.

Satellite imagery was first used to image marine mammals in 2002. Abileah, (2002), tested the Ikonos-2 satellite with a resolution of 82 cm on an orca (Orcinus orca) at a known location. These images showed potential, but the resolution was too low to have confidence in the detections. Since the Ikonos satellite launched in 1999 improvements in the spatial resolution of panchromatic satellite imagery have increased substantially from 82 cm to 31 cm in Maxar’s WorldView-3 satellite launched in 2014. This increase in spatial resolution allowed for increased applications of using satellites as a completely non-invasive and non-disruptive method for detecting marine mammals. Satellite images with very high spatial resolution are becoming more widely used for the detection of animals. Resolutions up to 31 cm and HD sharpening of images up to 15 cm have allowed for increased applications in studies of marine mammals (see below list of references).

Satellite imagery also has the benefit of being able to remotely image areas within a short timeframe from a remote location. Detection studies of large whales include fin (Balaenoptera physalus), humpback (Megaptera novaeangliae), southern right (Eubalaena australis), and gray (Eschrichtius robustus) whales. In high resolution satellite imagery probable southern right whales, fin whales, humpback whales, southern right whales, and gray whales were detected, described, and counted. More recently, medium sized whales including beluga whales (Delphinapterus leucas) and narwhal (Monodon monoceros) have been detected in remote areas in the Arctic. Furthering the detection of whales, satellite imagery has also been used for the estimating the abundance of beluga whales in the Canadian and Russian Arctic (please see: Fretwell et al., 2023; Sherbo et al., 2024; Watt et al., 2023). Satellite imagery is beneficial because of its completely remote and non-invasive nature, satellites are always circling the Earth and beluga whales appear to be blissfully unaware.

Drones are relatively non invasive and provide high resolution images or videos that can be used for monitoring marine mammals for behavioural and population size information. Resolution of drone imagery is 10x higher than captured satellite imagery (3 cm vs. 31 cm). I use the phrase relatively non invasive because if drones are flown at accepted heights whale disturbance is not observed. Recommendations advise that researchers fly drones above 25 m of belugas and approach larger groups with caution to minimize disturbance. Drone videos of belugas can be used to observe unique behaviour or group structure that could not be clearly seen from a boat (i.e. side vs. top-down view). Drone videos can be analyzed manually or using models to track behaviours. In addition to behavioural information, photo identification of wildlife can identify individuals and contribute to abundance estimates. Specifically monitoring beluga whale abundance using drones was conducted by creating a photo identification catalogue of beluga whales in Cumberland Sound by Ryan et al., (2022) and mark recapture methods were used for abundance. Photo identification was successful in this area as beluga whales with unique scars were able to be re-identified. Other locations where beluga whales do not have unique marks or scars may not be as successful with re-identification of individuals beluga whales. Ryan et al., (2022) found using drones was more successful to the boat-based photographs collected as the whales could be photographed closer and the images were higher in quality for capturing information.

Remote sensing imagery collection has limitations and it can be logistically challenging to capture clear images of large areas as environmental variables such as Beaufort Wind Scale, cloud, and sun glare can impact detection. However, this is also a limitation with traditional aerial surveys which can take weeks due to inclement weather (i.e., high winds or low cloud cover). Additionally, the collection of drone imagery and video is not completely remote unlike satellite imagery collection.

Overall, remote sensing methods including drones and very high-resolution satellites are highly valuable methods for monitoring behaviour and estimating beluga whale population abundance. In the future, as newer technologies become available imagery resolution will likely improve, thus increasing its applications. While imagery collected can require substantial time for processing and interpretation if done manually, there are more autodetection alternatives being developed. We anticipate faster and more accurate methods for reading images for the detection of beluga whales by using artificial intelligence such as pixel-based algorithms or semi-automated object-based image. Thus, we anticipate remote sensing methods to become more widely used for marine mammal population methods in the future.