Update in our Synthetic Aperture Radar (SAR) data pipeline

Update on SAR vessel detection and fixed infrastructure data pipeline

May 29: This post was updated after our previous announcement regarding the transition to an updated ocean-land mask for our Synthetic Aperture Radar (SAR) data processing. 

Overview of the change
Starting March 28, 2026, we transitioned to an updated “ocean-land mask” for processing SAR satellite imagery. This mask is a filter that allows our algorithms to “crop” the ocean and ignore landmasses, ensuring we only detect objects in navigable waters.

This transition was necessitated by a technical update in our cloud assets. To ensure the continuity of the platform, our team has implemented a new version of our masking tool.

This issue impacted these datasets in our MapAPIs and Data Download Portal:

    • SAR vessel detections
    • Sentinel 1 and Sentinel 2 fixed infrastructure

Following a comprehensive quality assurance (QA) period, we are sharing the results of our impact assessment to provide full transparency on how these changes affect our global vessel and infrastructure datasets.

Assessment period: March 13–26, 2026

Methodology: Detection percentages and coverage changes are calculated relative to the previous dataset totals to identify net gains and losses in the data.

1. Coverage
The mask update produces small coverage gains in some areas and small losses in others.

    • Coverage Gain: The new mask adds approximately 10,000–16,000 km² per day to our processing pipeline (roughly 0.15–0.25% of total global coverage).
    • Net Benefit: The area added in previously “masked out” zones outweighs the minimal area lost during the transition.

2. Vessel Detections: net gains and specific losses
The updated mask allows us to detect closer to the shore in some areas, while also removing some legacy detections in specific nearshore environments.

    • Gains (2.1–2.9%): New detections have emerged in nearshore areas that were previously excluded with the old  mask. Coverage gains slightly exceed losses across all continents.
    • Losses (1.7–2.1%): Some existing detections were dropped, primarily in nearshore areas no longer covered by the new ocean mask.
    • Top Affected EEZs: Norway, Indonesia, China, Turkey, Japan and Chile saw the highest volume of change. Notably, South America and Australia experienced higher loss rates (2–5%) compared to the global average.
    • Regional Focus: Europe and Asia account for the majority of changes, while Africa remains the least affected continent (<2% change across all categories).

3. Fixed Infrastructure Impacts
In this early analysis, the mask update has affected our ability to detect certain offshore structures. We analyzed which ones fall outside the new ocean mask.

  • Infrastructure Displacement:   .
    • At the time of the mask update, there were 427 structures that ended outside the new mask.  
    • With the new mask we lost the ability to track those infrastructures, so by May 29 some of them no longer exist and we also have 66 ongoing structures that were affected.  
  • Industry Breakdown: The impact is dominated by oil infrastructure.
  • Geographic Distribution: Asia is the most affected region (214 structures), followed by South America (63), North America (56), and Europe (50).
  • EEZs with Highest Impact: China (83 structures), USA (38), Chile (32), and Indonesia (23).
  • ⚠️ Our team is hard at work analyzing the recent performance gains on our fixed infrastructure. We’re also diving into some exciting new methodologies to improve our approach and will have a full update to share with you very soon.

Next Steps

  • Data Availability: 
    • 🟢 SAR Vessel Detections: All Map, API, and Data Download Portal data are now reflecting these updates.
    • 🔴 Fixed Infrastructure: we stopped the pipeline since March 24, 2026, which means:
      • Map and API: While infrastructure data remains visible on the Map and API, please note these records reflect our last processing run from March 24, 2026. Because these detections have not been assigned an ‘end date,’ they will remain visible in the dataset as ‘active’ until the pipeline resumes.
      • Data Download Portal: We have pending to generate Jan 2026 data in April because of this issue.
      • We continue evaluating the results to enable the data processing as soon as possible. 

Questions?
We are committed to providing the most accurate open-access data available. For technical inquiries regarding specific datasets or how this might affect your work, please reach out to our team at [email protected] .

Related Updates

Scroll to Top