South Wight Marine Protected Area

Intelligence Report – July 2022

South Wight is a complex and busy multi-use marine protected area.

General Key Finding

The South Wight MPA experiences persistent high levels of small-scale vessel traffic.

Vessel Tracking Key Finding

Machine learning and AIS analysis identified several vessels of risk, with AIS detecting vessels operating in and around the MPA.

Remote Sensing Key Finding

Remote sensing indicated the South Wight MPA is at moderate risk of suspected non-compliance during the period of analysis due to the large number of low and high-risk detections.

Remote sensing synthetic aperture radar (SAR) was able to identify targets as small as 3m in the MPA. From this a good understanding of the site use can be deduced; complex, varied and heavily trafficked.

Observed Activity

  
      AIS Analysis Alerts:
      10 (3 vessels)
    
      Machine Learning Alerts:
      33 (20 vessels)
    
      SAR Images:
      8
    
      EO Images:
      3
    
      High Risks:
      0

The map to the right shows all the high and low risk ‘dark vessel’ detections in and around the South Wight MPA from SAR swaths and EO images taken between April and July 2022.

An aggregation of detections was located on the western side of the Isle of Wight and the Needles, and there was an even distribution of detections across the east coast of the Isle of Wight.

These detections highlight the difficulty in carrying out patrols across the entire MPA.

The Electro-optical image below shows an example of the type of pleasure traffic operating within the MPA. These jet skis, approx. 2m in length and their characteristic wake were visible with high resolution colour EO imagery.

Conclusions

SAR identified a moderate level of dark vessel detections but was limited to determining presence and could not be used to determine vessel type or activity.

Electro-Optical high-resolution imagery, though limited in its coverage, could be analysed to determine vessel type and potential activity with a high degree of confidence.

Vessel tracking enabled an understanding of vessel behaviour and broadly showed good compliance. When the national rollout of iVMS has been completed, machine learning could provide insights and efficiencies for monitoring compliance on this site.

Recommendations

Intelligence collected through remote sensing could be used to determine behaviour patterns and identify vessels and subsequently inform patrol planning and direct patrol resources. Using machine learning to analyse vessel tracks could provide efficiencies in monitoring compliance within bylaw areas, particularly when iVMS is in use within the district.

Extended Findings

Over the monitoring period there were 3 high risk detections identified within the MPA. Remote sensing imagery identified that the highest concentration of at risk detections was off the southeast side of the Isle of Wight, close into the shore. Therefore, most potential risks in the MPA would be identifiable from shore-based monitoring.

The high-resolution SAR was capable of picking up very small targets and the high-risk detections ranged from 5 – 63 m in length.

Several clusters of detections were identified. One cluster which could be indicative of dark vessel activity was located west of the Needles around an area of bathymetry indicative of a good fishing area. Two further clusters close to the shore off Sandown and Shanklin could align with anchored vessels or fishing vessels operating close inshore.

AIS tracking sources identified 2 high risk vessels with possible fishing activities in the MPA during the monitoring period.

Machine learning did not identify any risks within the South Wight MPA; iVMS is not yet operational in this area and vessels over 12 m transmitting on VMS are unlikely to operate in this area.

EO and vessel tracking analysis identified a large presence of pleasure vessels in the MPA, comprised of high-speed craft, sailing yachts and sport fishing traffic showing a high persistence.

Risk Assessment

From remote sensing detections identified, the main risks for the South Wight MPA originate from vessels between 13 and 63 m in size. Based on the number of detections identified in remote sensing, the primary risk area was the nearshore area. From analysis of SAR swaths taken over a short time period, the persistence of detections operating in the same area was low. This indicated that the main risks were opportunistic vessels operating over short durations, these vessels are likely to be below 12m and therefore do not have a VMS requirement. Furthermore, the low reporting rate of vessels with VMS may allow these vessels to trawl within the site. The majority of vessels identified on tracking data were small vessels with class B AIS transponders. AIS identified a risk of trawlers encroaching on the edge of the MPA both in the western trawl corridor and in the eastern MPA.

Next Steps

Overall, from remote sensing sources, a low risk of likely unauthorised vessel activity was identified over the South Wight MPA. Vessel tracking and EO data sources identified a low risk from unauthorised vessels, and SAR imagery detected a moderate persistence of dark vessels across the MPA. Detections were focused close to shore, and this could be indicative of possible dark fishing vessels or pleasure traffic, operating without AIS.

SAR is capable of picking out vessels of different sizes, and is well suited for describing vessel presence, but not type or activity. Due to the relatively small size of the site SAR would not be useful as a patrol support tool, as once a vessel has committed to patrol the area, it is likely to cover the whole site rather than target a specific location within the site. SAR is however useful for establishing a general picture of activity, strategic planning and identifying high risk areas and times.

A consideration for the site would be to install coastal surveillance radar (CSR). This would provide a picture of presence/absence of all vessels operating around the Island and would be an economical way to collect the same information gathered through SAR analysis with a shorter lead time and 24/7 monitoring.

EO provided high resolution imagery of detections and when available proved a valuable tool to characterise vessel types and possible behaviour. It is recommended to use EO to supplement additional remote sensing and monitoring sources.

Based on the risks identified in the project it is recommended to apply machine learning to iVMS data upon completion of the national rollout, in order to efficiently and effectively identify any non-complaint activity which occurs within the restricted byelaw areas.