| Summary: | New tools and methods are needed to monitor marine resources so that industrial activities can be conducted without (or at least minimizing) adverse impacts on species of concern. UAVs can be used as a complementary or alternative tool to current methods for monitoring and research of marine mammals. This study highlights current knowledge gaps and the need for further empirical testing of these systems. The capabilities of a system must be well understood before field trials are carried out. Platforms and sensors have different qualities and limitations, and will perform differently depending on the type of monitoring needed. When conducting field tests, it is important to acknowledge the many factors that may bias image analyses. Factors external to the survey equipment (such as environmental features) may affect UAV data differently than visual observer-based aerial survey data (hereby manned-surveys). Changes in pixel size due to aircraft movement may affect the resolution in which an animal is present within an image and is therefore a measure that should be included in analyses of digital imagery. Certainty of detections is a measure of relevance for such analyses as it provides a better understanding of the effects of environmental and survey-related covariates on image analysts' capabilities to detect an animal. Multiple aircraft or single aircraft maneuvers are often conducted to validate observations and estimate animal availability. To increase the number of detections when using multiple aircraft, one must consider animal availability parameters that can bias estimates of abundance or density. Simulation studies considering survey features and animal behavior can be used to improve data acquisition using digital imagery (e.g., deployed by UAVs). Overall, this highlights the complexity of monitoring programs, and shows how technological progress is valuable not only for environmental scientists, but also for industry managers and regulators.
|
|---|