Proceedings of the Institute of Acoustics

 

Noise disturbance – baseline level monitoring in the Solent

 

Najwa Adnan-Smith, Tetra Tech
Dawit Abraham, Tetra Tech

Acknowledgements Natural England

 

1 INTRODUCTION

 

The Solent coastline hosts thriving harbours, ports and other coastal industry, meaning there is a high volume of activity interacting with the marine environment. The majority of this activity creates a level of noise that can lead to the disturbance and displacement of Special Protection Area (SPA) bird features. The SPAs in the Solent area with marine components are the Solent and Southampton Water SPA, Portsmouth Harbour SPA, Chichester and Langstone Harbours SPA and Solent and Dorset Coast SPA.

 

In particular, most birds are sensitive to above water noise. This pressure relates to any anthropogenic loud noise made onshore or offshore by construction, vehicles, vessels, tourism, mining, blasting etc. that may disturb birds and reduce time spent in feeding, resting or breeding areas.

 

Above water noise is benchmarked as the introduction of airborne noise above background levels, however there is currently a lack of data on what background noise levels are. Therefore, assessing the risk of an activity disturbing birds through the introduction of above water noise has proven difficult.

 

This work, commissioned and funded by Natural England, has been designed to fill this data gap for the overwintering period through noise monitoring in key areas across SPA sites in the Solent with the objective of providing data on background noise levels in order to more accurately determine the likely significant effect on birds when responding to anthropogenic noise.

Nine areas of high activity have been identified. These areas are hotspots for anthropogenic activity and cross over with key areas of designated SPAs. These areas are therefore used as the main sites for the background noise surveys and are representative of the Solent as a whole.

 

2 METHODOLOGY

 

2.1 Long-term Baseline Noise Monitoring

 

A monitoring survey was undertaken to characterise the baseline ambient noise levels currently experienced on site at nine SPA locations. Measurements were taken in general accordance with BS 7445-1:2003: Description and Measurement of Environmental Noise – Guide to quantities and procedures¹.

 

The baseline monitoring survey was undertaken monthly from October 2023 to February 2024 at the nine SPA locations presented in Figure 2.1 with measurements being made unattended in 5-minute intervals over a minimum of 72-hour period at each site per month.

 

Figure 2.1: Long-term Unattended Noise Monitoring Locations

 

 

2.1.1 Data Analysis for Long-term Noise Monitoring

 

The collected data per measurement period were divided into daytime hours (07:00 – 23:00) and night-time hours (23:00 – 07:00) to reflect the operational hours of anthropogenic activities. The local weather conditions for the duration of measurement were established using Weather Underground, an online resource which provides real-time and historical weather information. Where appropriate due to periods of heavy rain or high wind speeds, data was omitted from analysis.

 

Background noise levels are usually described using the L A90 index (i.e. the sound level exceeded for 90% of the measurement period). This sound index was chosen to represent the background noise levels in the areas studied and the modal L_A90 sound level of each 5-minute measurement is used to represent the overall background noise levels during the daytime and night-time periods.

 

It should be noted that some measurement periods were omitted due to equipment failure as a result of poor weather conditions or unsuitable weather conditions to set up the equipment. All locations were monitored for a period of at least four months except for 2Hythe and 6Ryde due to meter failures during both October and November long-term surveys at 6Ryde and during February long-term survey at 2Hythe.

 

2.2 Short-term Noise Monitoring

 

Short-term noise monitoring was undertaken to coincide with the long-term unattended baseline noise monitoring in order to observe the bird species present, their behaviour and any responses to anthropogenic noise in the area. Observations were made during high, low, rising and falling tides.

 

Up to three observation positions near to the long-term noise meter were chosen to maximise the likelihood of observing bird species. Bird count and species identification within an approximate range of 500m were undertaken using binoculars and a telescope for a total of three times during the observation period.

The short-term noise meter was set up to measure the sound levels in 1/3 octave bands and 1-second intervals over a minimum period of 2 hours during the daylight hours. Observations of any anthropogenic noise were noted to include the source of noise, time of the noise event, whether the disturbance is also visual in nature, estimated sound pressure level at the measurement position, and the distance of noise source to the birds being observed as well as the measurement position. The respective distances between the noise source and the location of the bird and the measurement position are estimated using the map on the Survey123 application used. Bird responses to the anthropogenic noise were categorised as one of the following:

0 – no response

1 – freeze/stress response

2 – staying at site but moving away from noise

3 – flight response with settlement within 100m

4 – flight response with settlement beyond 100m

 

2.2.1 Data Analysis for Short-term Noise Monitoring

 

Where any bird response was noted during the survey, the sound pressure level at the meter location is used to estimate the sound pressure level at the location of the bird using the relationship below as referenced in The Little Red Book of Acoustics written by R Watson and O Downey².

 

 

L_(n) is the sound pressure level at a location and r_(n) is the distance from the noise source to the location. In this case, ‘1’ refers to the monitoring location and ‘2’ refers to the location of the bird.

 

Similar to the above, the loudest instantaneous noise levels L_AFmax which correspond to the noise events where bird responses were observed are also estimated at the location of the bird.

 

3 RESULTS

 

3.1 Background Noise Levels

 

The background noise levels during the daytime vary from month to month by up to 5.0 dB(A) at all locations except for 5Thorness Bay, 6Ryde and 7Portchester where the differences are in the range of 7.0 to 13.0 dB(A).

 

For the night-time, the background noise levels vary from month to month by up to 3.0 dB(A) at locations 1Lymington, 3River Itchen, 4Hook Lake, 9Emsworth whilst the other locations vary by up to 10.0 dB(A).

 

Figure 3.1 below presents the overall average daytime and night-time background noise levels across the entire monitoring period for each location.

 

Figure 3.1: Overall Average Background Noise Levels Across the Whole Monitoring Period

 

 

Generally, the overall average daytime background noise levels at all locations are between 43.0 dB(A) to 49.0 dB(A) with the exception of location 8Farlington Marshes where the overall average daytime background noise level is 69.0 dB(A).

Similarly, the overall average night-time background noise levels generally range from 36.0 dB(A) to 44.0 dB(A) at all locations except 8Farlington Marshes where the overall average night-time background noise level is 52.0 dB(A).

 

3.2 Bird Response

 

Where there were bird responses observed during the short-term noise monitoring survey, the noise events which triggered the response are analysed. The analysis includes the type of response from the birds as well as the estimated sound pressure level (SPL) and L_Amax levels caused by the noise event at the location of the bird.

 

Bird responses were only observed at locations 2Hythe, 3River Itchen, 6Ryde, 7Portchester and 9Emsworth with SPL from the noise events at the location of the birds estimated to be in the range of 46.9 – 85.4 dB(A). These bird responses were from 11 species out of the 51 species observed throughout the attended short-term monitoring. Most of the responses observed were also from Brent Goose (BG). The type of response observed were mainly of flight response with settlement within/beyond 100m with a few freeze/stress responses.

 

Although loud noise events (SPL between 30.0 dB(A) to 79.7 dB(A) at the location of the birds) caused by anthropogenic activities were observed at the other locations, no bird responses were noted.

 

Furthermore, the noise events which triggered bird responses include horns from vehicles, airplanes and helicopters passing overhead, trains passing, metal works, industrial noise, boats and hovercrafts, and people walking/talking. These are mostly also visual in nature. It should also be noted that all of the observed bird responses at the location 6Ryde were triggered by frequent hovercrafts.

 

The range of estimated L_Amax levels at the location of the birds for the noise events which resulted in bird responses is between 48.2 – 86.4 dB(A). However, it should be noted that no bird responses were observed for other perceptible noise events with similar estimated L_Amax levels in the range of 32.2 – 81.2 dB(A) at the location of the birds.

 

The estimated L_Amax levels which triggered bird responses are similar (within + 0.5 to 4.1 dB(A)) to their corresponding estimated SPL.

 

4 DISCUSSION

 

4.1 Background Noise Levels

 

The results of the long-term noise monitoring show that the overall average background noise levels at all locations are between 43.0 dB(A) to 49.0 dB(A) during the daytime and 36.0 dB(A) to 44.0 dB(A) during the night-time. However, higher overall average background noise levels (69.0 dB(A) daytime and 52.0 dB(A) night-time) at 8Farlington Marshes were recorded.

 

The long-term noise monitoring location LT8 at 8Farlington Marshes is approximately 120m south of the major road, A27. Based on strategic noise mapping data for road sources published by Extrium, the noise levels from road traffic at the monitoring location are predicted to be between 65.0 – 69.9 dB L_Aeq,16hours during the daytime and between 60.0 – 64.9 dB L_Aeq,8hours during the night-time. Other monitoring locations are at least 500m from any major roads. In the case of LT6 in 6Ryde, the main road A3055 (approximately 110m south of LT6) is considerably less noisy than the road A27 adjacent to LT8. As such, the noise contribution from road traffic is highly likely to be the reason for the higher background noise levels measured at LT8.

 

Despite higher background noise levels, at least 29 bird species, both breeding and non-breeding, were observed in the area during attended monitoring in 8Farlington Marshes. However, further and/or longer monitoring is likely required to determine if the birds are affected by the anthropogenic noise in the area.

 

4.2 Comparison of Observed Noise Events to Background Noise Levels

 

The results presented in this report indicate that bird responses typically occur when the sound pressure level at the location of the birds is at least 20.0 dB(A) higher than the typical background noise level L_{A90,16hours(daytime)}. This is relatively comparable to the study on laying hens by J. L. Campo, M. G. Gil and S. G. Dávila (2005)³ which showed that hens were found to be more stressed and fearful when exposed to higher sound levels (90 dB) for 60 minutes which consisted of background noises plus truck, train and aircraft noises compared to the control group which was exposed to only background noise levels at 65 dB.

 

However, this is unlikely to be the main factor for the bird responses observed during the attended noise monitoring as several of the bird responses noted at the location 9Emsworth were triggered by people walking into the beach where the sound pressure levels at the locations of the birds are estimated to be only 3.9 dB(A) to 11.3 dB(A) above the daytime background noise level. This response is likely to have been triggered by the visual nature of the disturbance rather than noise. It was also observed that majority of the noise events with sound pressure levels less than 20 dB(A) above the background did not trigger bird responses.

 

J. R. Barber et al. (2009)⁴ suggested that animal responses to anthropogenic noise are likely to depend on the intensity of perceived threats rather than on the intensity or level of noise. This may apply to most of the observations made during the attended surveys but particularly in the case of the bird responses observed from people walking into the beach at 9Emsworth.

 

Furthermore, the bird responses observed at 9Emsworth from airplane passing overhead and at 7Portchester from a boat leaving the harbour showed a difference between the sound pressure levels and background noise levels of only 9.0 dB(A) and 14.3 dB(A) respectively.

 

In a study of brent geese and human disturbance, Owens (1977)⁵ suggested that larger birds with slow wingbeats such as Great Black-backed Gulls (GB) are also liable to causing flight responses in brent geese and intensity of responses to aircraft may be partly due to the visual resemblance of aircrafts to large birds. This could be the trigger to the bird response observed at 9Emsworth mentioned above.

 

At 6Ryde, three noise events from hovercrafts which resulted in a difference of more than 20.0 dB(A) between sound pressure levels at the locations of the birds and the background noise level showed no response. However, bird responses were observed for subsequent noise events from hovercrafts within the same survey period.

 

Most of the bird responses observed particularly at 6Ryde were also from brent geese. This might suggest that brent geese are more sensitive to noise events, but it is more likely that the brent geese were loafing on the water closer to the noise sources such as hovercrafts compared to other species including waders which were foraging along the shoreline.

 

Furthermore, the type of responses observed were mainly flight responses with two freeze responses due to airplanes overhead. It is highly likely that other freeze responses were not immediately noticeable compared to the flight responses and thus were missed out. It is nearly impossible for the surveyor alone to analyse whether a bird is showing a freeze response to a noise event when simultaneously observing other birds of different species in the survey area. As such, for any future works, it may be beneficial to also record the birds and retrospectively analyse the data by playing back the recording.

 

4.3 1/3 Octave Band Analysis of Noise Events

 

In a study on the effects of highway and urban noise on birds conducted by R. J. Dooling et al. (2019)⁶, it is suggested that anthropogenic noise can affect birds’ abilities to detect prey, assess their acoustic environments and communicate with other birds. If the noise includes enough energy in the bird’s region of best hearing or dominant frequency, at close distances, the noise can have a significant impact on how well the birds can hear their species-specific vocalisations. This in turn may cause behavioural and/or physiological responses from the birds.

 

This is also shown in a study by Rheindt (2003)⁷ which consisted of population assessments in an oak-beech forest close to a motorway where it was concluded that bird species with higher-pitched vocalisations or songs with dominant frequencies well above the typical frequencies of traffic noise (up to 1 kHz) were less susceptible to noise pollution. Rheindt also stated that most bird vocalisations, in contrast, are in the range of 2 kHz to 9 kHz.

 

The 1/3 octave frequency data for each noise event which triggered a bird response shows no obvious correlation between specific frequencies and bird response. However, the sound pressure level of the noise event at each frequency is generally above the background noise level measured during the attended short-term monitoring. Figure 4.1 below present the visualisation of the 1/3 octave frequency data (from 12.5 Hz to 20 kHz) for all noise events which triggered bird responses.

 

Figure 4.1: 1/3 Octave Frequency Leq (top) and Lmax (bottom) for All Noise Events with Bird Responses (dB linear)

 

 

* The percentage of the birds disturbed is the percentage calculated from the total number of the same species observed during the survey.

 

[The common names of the bird codes used above are given below.

Breeding birds: coot (CO), grey heron (H.), herring gull (HG), mute swan (MS), black headed gull (BH), oystercatcher (OC)

Non-breeding birds: curlew (CU), dunlin (DN), brent goose (BG), redshank (RK), turnstone (TT)]

 

The events above are grouped by the species triggered as responses that are frequency dependent would also likely be species dependent due to vocalisations being species-specific. However, no obvious correlation can be seen between specific frequencies to bird responses observed during the short-term noise monitoring and majority of the noise events have low dominant frequencies (below 200 Hz).

 

As such, it is unlikely that these bird responses were caused by any specific frequencies, particularly as most bird vocalisations, and their dominant frequencies, are in the much higher frequency range.

 

5 CONCLUSION

 

Typically, the daytime background noise levels range between 43.0 dB(A) to 49.0 dB(A) at all monitoring locations with the exception of one location – Farlington Marshes within the area of Chichester and Langstone Harbours SPA – where the daytime background noise level is 69.0 dB(A) due to the location’s proximity to a major road.

 

In addition to this, short-term attended noise monitoring was undertaken to coincide with the long-term monitoring. During the short-term noise monitoring, observations of anthropogenic noise and any bird responses as a result of the noise were made. The sound pressure levels which triggered the bird responses were estimated in order to gain some understanding of the reason for the response.

 

The results show that birds are more likely to respond to noise disturbance when the sound pressure levels at the location of the birds are at least 20.0 dB(A) above the typical background noise level. However, the visual nature of any noise disturbance is also likely to cause responses from the birds. The findings of this study will help to determine the impacts of anthropogenic noise on overwintering birds in the Solent; a key challenge given the national and international significance of these populations.

 

6 REFERENCES

 

1. British Standards Institution (2003) BS 7445-1:2003: Description and Measurement of Environmental Noise – Part 1: Guide to quantities and procedures. London: British Standards Institution.
2. Watson, R. and Downey, O. (2008) The Little Red Book of Acoustics: A Practical Guide. Second Edition. Blue Tree Acoustics.
3. Campo, J.L., Gil, M.G., & Dávila, S.G. (2005) Effects of specific noise and music stimuli on stress and fear levels of laying hens of several breeds. Applied Animal Behaviour Science 91:75-84.
4. Barber, J. R., Crooks, K. R., Fristrup, K. M. (2009) The costs of chronic noise exposure for terrestrial organisms. Trends in Ecology and Evolution 25:180-189.
5. Owens, N. W. (1977) Responses of wintering brent geese to human disturbance. Wildfowl 28:5-14.
6. Dooling, R. J., Buehler, D., Leek, M. R., Popper, A. N. (2019) The impact of urban and traffic noise on birds. Acoustics Today 15:19-27.
7. Rheindt, F. E. (2003) The impact of roads on birds: Does song frequency play a role in determining susceptibility to noise pollution. Journal of Ornithology 144:295-306.