The consideration of flight heights is a key factor determining how seabirds interact with offshore wind farms. Of particular interest is the need to accurately predict flight heights, in order to feed into predictive Collision Risk Models (CRM). Data within CRM have traditionally relied quite heavily on boat surveys collected by observers, but now routinely, different methods are being used to estimate flight heights of birds.

To date, the relative suitability of methods for the collection of flight heights has not been rigorously appraised. For marine birds, this has primarily included visual methods from boat-based surveys, but more recently remote monitoring techniques such as high definition aerial imagery (images, video and spectrographic techniques) have been used. Alternative methods have also included bird-borne telemetry, radar, laser rangefinders, thermal imagery and acoustic methods, but their suitability to obtain reliable flight height distributions for potential use within CRM has not yet been assessed. This report details and compares each method for estimating flight heights, evaluating the relative merits and disadvantages of each method.

We conducted a literature review and assessed each method based on their (1) general use in deriving flight height distributions of bird species, (2) the testing, calibration and validation of these methods that have been carried out, and (3) the relative advantages and disadvantages of each. We then collated available flight height information as a first step in comparing results from different survey platforms for different species.

Boat-based surveys using observers have visually assessed the flight heights of individual birds at sea, with observers placing observations into flight height bands. These methods have been widely used to give flight height distributions. However, they can only be used in the day in good weather conditions and have a degree of imprecision, as they do not ascribe a specific flight height to each observation. Issues of disturbance/attraction are also present, with some species being sensitive to presence of the survey vessel and others attracted to it. A further concern lies in the safety implications of sending surveyors to sea, especially for offshore sites.

More recently, aerial high definition methods have been used to estimate flight altitudes of birds from visual stills and video methods. Such methods provide a permanent record that can be revisited, re-analysed and quality assured, as well as avoid disturbance issues if the aircraft is flown at a suitable altitude (above 460m). Survey conditions must be suitable to allow identification (avoiding high winds or very rough sea state) and good visibility (avoiding foggy conditions) and is generally restricted to daytime operations.

Spectrometry methods are under development which uses two high resolution cameras to create a three-dimensional image from which flight height can be accurate determined, and these methods can be used throughout day and night. Further testing and validation of these methods is ongoing and is the subject of trials in the UK and overseas.

Telemetry methods include altimeter and geographical positioning system (GPS) derived altitude data and although relatively few studies have investigated their use, they have recently been used to estimate bird flight heights. Studies using altimeters have reported small errors on estimates. Initially, altimeters were too heavy for some species and could not be deployed simultaneously alongside other sensors recording latitude and longitude to obtain geo-referenced three-dimensional information. However, this issue is now being overcome through advancements in technology. In contrast, GPS flight altitudes have higher error surrounding 8 measurements, e.g. 15-20 m and accuracy bias, but modelling approaches can be used to obtain flight height distributions, and 3-D behaviour can be investigated in great detail.

Telemetry may cover the full range of habitats used by birds, allowing more in depth analysis of flight height distributions. Telemetry can also produce flight height distributions for particular protected sites. Telemetry must be assessed for potential impacts to individual birds on which devices are mounted, and has considerably lower sample sizes than radar or other techniques. They are often deployed on particular life stages (e.g. breeding birds, rather than immatures/non-breeders, as they return to colony frequently), and for some species data collection may be restricted to the breeding season only (i.e. when using GPS and GPS+altimeter methods). Typically, harness methods are required for study during the non-breeding season, which may not be suitable for smaller species. The impact of a telemetric device and attachment method on an individual’s behaviour needs evaluation in any study carried out. The population-level representativeness of the telemetry data collected using a relatively small sample of birds also requires consideration.

Radar and other methods (laser rangefinders, infrared imagery, and sound) have been widely used to estimate flight heights and can have a high level of accuracy. Radar can give a greater temporal coverage being used in all weather conditions and times of the day/year. Radar can be more restricted in spatial coverage, although it may still be sufficient to understand questions posed by offshore wind farm impact assessments (e.g. at a specific wind farm). Radar sometimes cannot identify very low altitudes due to clutter. There is also a significant problem with species-level identification or even identifying groups from single individuals if radar techniques are used in isolation. As such, it may be necessary to use radar in combination with an additional method.

Validation comparisons of the flight heights obtained from radar and laser rangefinders have been carried out, finding good agreement in many instances. Furthermore, altimeters have been compared to altimeters on planes to calibrate instruments. Boat-based visual flight height distributions have been compared to GPS flight height distributions of Lesser Black-backed Gulls finding good agreement during the daytime. However, laser rangefinders have also recorded higher proportion of lower altitude measurements than radar in other studies; for example due to difficulties in selecting and recording more precise estimates for bird further away. Moreover, visual methods are restricted towards daytime use (or night-time for ceilometers and moon-watching methods) and given species have been found to differ in flight behaviour in day and night, which may bias the overall flight height distribution.

The use of telemetry and radar in combination may offer advantages over traditional visual methods, including wider spatial (and vertical) coverage. Telemetry in particular has the potential to deliver data on individuals’ behaviour, but to our knowledge no studies have yet compared data collected alongside different telemetry sensors nor to other methods such as radar. Radar cannot identify individuals to species level, requiring more targeted validation from other methods, such as visual observations.

Other methods reviewed included laser rangefinders, inclinometers, acoustic methods, thermal imagery, moon-watching and ceilometers and ornithodolites. These methods were considered useful but generally have certain restrictions that limit their use to a more supplementary role or are not suitable for use in the marine environment. These limitations include restrictions to lower altitudes, day or night biases in measurement period, and smaller spatial scale coverage. Laser rangefinders, inclinometers and ornithodolites in particular work best with a stable platform to reliably lock on to and register flight altitudes of targets, making them less suitable for use on unstable platforms in the marine environment.

The site-specific practicalities of different technologies are ultimately likely to determine the method used for a given study. Therefore, no single method is recommended for all situations. However, the most feasible methods reviewed here to provide reliable estimation of bird flight altitudes offshore are: high definition digital imagery surveys (such as aerial surveys), telemetry methods (altimeter and/or GPS) and X-band radar. A combination of these methods (deployed across multiple seasons, sites and species) is likely to give the most accurate and widely applicable information on seabird flight heights. This could also be supported by other methods such as visual, laser rangefinders and infra-red thermal methods, where applicable for validation/ground-truthing.