Abstract
This study was jointly initiated by NIRAS, Aarhus University and Energinet. The study was independently carried out by NIRAS and Aarhus University and funded by Energinet. The objective of the study was to investigate whether the presence of active geophysical survey ships using USBL (Ultra-Short BaseLine) acoustic positioning systems had any impact on the baseline data regarding the presence of harbour porpoises in the North Sea I survey area. Geophysical surveys utilize USBL systems to keep track of their underwater equipment. USBL systems emit signals at frequencies and source levels known to cause displacement of harbour porpoises from the area. A previous study found modelled impact ranges of up to 3.0 km. If an effect was found, and the baseline data thus could not be said to represent the “natural” unaffected situation, this would need to be accounted for. Con- sequently, the purpose was further to quantify the effect and to develop a method to compensate for the impact on the baseline data.
A study examining effects of USBLs on presence of harbour porpoises has not previously been carried out and different approaches were therefore tested in this report to find the most optimal method to possible correct for the impact on USBL on baseline data. Data was collected with 42 F-PODs (porpoise loggers) and 6 broadband acoustic recorders (SoundTraps) from April 2023 to November 2023. Three of the broadband stations were lo- cated inside the geophysical survey area (impact stations) and three were located outside (control stations). Time- stamped GPS tracks of the geophysical survey vessels were obtained for the same time period from the geophys- ical survey operators. In the broadband data, USBL pulses were found and distance to nearest survey vessel cal- culated. For each pulse the source level was back-calculated and from that the potential impact range was calcu- lated based on the harbour porpoise behaviour criterion Lp,rms,125ms,VHF = 103 𝑑𝐵 𝑟𝑒. 1𝜇𝑃𝑎. Predicted impact ranges between 1 km and 5.5 km were observed.
During the analysis, it became evident that USBL and USBL-like signals were not solely emitted by geophysical survey vessels. A significant portion of these signals was assigned to unidentified sources. It was found that some of these unknown vessels were likely trawlers, which utilize various USBL systems to monitor their trawls and measure their catch.
From the F-POD data, three indices of harbour porpoise presence were calculated: porpoise positive minutes (PPM), clicks per minute (CPM) and waiting time (time between consecutive porpoise acoustic encounters) and the effects of USBL use and received level were estimated using mixed-effect statistical models.
The following hypothesis were tested:
A) PPM and CPM both correlate negatively with presence of USBL signals from geophysical surveys.
B) Waiting time from geophysical USBL signal to first harbour porpoise encounter correlates positively with received level of USBL signal from geophysical surveys.
Results show that PPM and CPM decreased with increasing received level of USBL pulses. Oppositely, waiting time increased with increasing received level. The statistical models showed that waiting time from USBL pulse to first harbour porpoise encounter on average increased to 196 minutes (95% confidence intervals: 154 - 239) as opposed to periods without USBL pulses where the average waiting time between consecutive harbour porpoise encounters was 66 minutes (95% confidence intervals: 31-102).
In answer to the main objective of this study, it is concluded that the baseline data collected for harbour porpoises in the North Sea I survey area was biased during the presence of geophysical survey vessels. To address this issue, different approaches for excluding affected baseline data were tested. Five scenarios were tested in an attempt to compensate for the impact of geophysical survey vessel presence, on F-POD detections, evaluated in effectiveness by examining change in CPM and PPM.
First a basic approach was tested, where all days with an active geophysical survey vessel using USBL was present within 3.5 km distance of an F-POD station, were excluded. The 3.5 km distance was chosen as an assumed impact range based on a previous study where approximately 3.0 km predicted impact range was found. Next, we tested an approach where, in addition to the basic approach, the day following a geophysical survey vessel presence within 3.5 km distance was also excluded to account for geophysical survey vessel presence near midnight. Since the analysis of USBL detections showed potential impact ranges of up to 5.5 km, the two approaches were also tested using an increased exclusion distance of 5.5 km, as well as an approach where we also removed the fol- lowing day. All four approaches showed an effect on PPM and CPM, however it was not consistent when analysed temporally and spatially over the three impact stations, indicating that both affected and unaffected data were excluded from the dataset.
Finally, waiting time was tested in an advanced approach, where the first 239 minutes (mean + 95% CI rounded to nearest minute) following geophysical survey vessel presence within 5.5 km distance of the F-POD stations, were removed from the impact stations. Hereafter, mean CPM and PPM per month was recalculated to test for effect on the entire dataset. The recalculated CPM and PPM for all tested F-POD datasets, consistently showed an increase in mean values, indicating that the approach had a compensating effect. Of the five approaches tested, the ”advanced” approach also required the lowest amount of data to be excluded from the dataset; 18.3% of minutes in the dataset for the three tested impact stations. To apply the advanced approach to the remaining F-POD stations, geophysical survey vessel presence within 5.5 km of each individual F-POD station must be mapped, after which the active presence time + 239 minutes can be excluded from the individual F-POD datasets.
In conclusion, the advanced approach was found to be the best available approach for correcting for the baseline data which was impacted by the USBL systems used by geophysical survey vessels within the North Sea I survey area. Impacted data were removed because they provide a false negative impression of the presence of harbour porpoises in the area, when the geophysical survey vessels were present. Hence, when removing impacted data, the intention is to provide a more natural picture of the presence of harbour porpoises in the North Sea 1 area. While the approach is considered valid for the North Sea I survey area, the findings cannot be directly applied to other areas, species or time periods, without project specific studies. It is difficult to generalize the validity of the advanced approach from this study to other cases, since this is the very first study trying to quantify the effect of USBL deterrence on harbour porpoise presence. It is therefore not possible to predict whether the increase in waiting time found here, will be similar in other areas.