Periodic Reporting for period 2 - ARCOS (Arctic Observatory for Copernicus SEA Service)
Período documentado: 2022-03-01 hasta 2023-09-30
Resulting from the dramatic decline in the ice cover in the Arctic Region, new economic opportunities are emerging, such as new traffic routes for the transport of goods from east to west and access to recently accessible natural resources. However, these changes have the potential to increase environmental threats and security concerns in the region coming from interest on the intensive exploitation of natural resources on the artic land areas (e.g. Gas resources); increasing risk of seaborne disaster and oil spills or establishment of permanent human facilities compromising European security. The consensus within the EU to maintain a multilateral cooperation approach to ensure stability and dialogued solution in the region triggers an increasing demand of situational awareness.
The objective of ARCOS is to design and implement an early-warning system providing continuous monitoring of the Arctic Region. Designed to generate actionable products in the security domain by processing and fusing multi-sensor data, the system integrates available information from space, non-space sources and products available from multiple Copernicus services.
ARCOS provides information services at three different levels of scale and user interaction:
• Level 1: Automatic Early-warning System. Integration of space and non-space data sources for the triggering of alarms on the region when certain conditions happens. Automatic early-warnings are generated in case anomalous behaviors are detected. For this wide-area monitoring, automatic extraction of analytics and AI techniques are applied.
• Level 2: User-Driven Alert System, where space and non-space data are processed on specific locations provided by the user. The alarms can be configured as contextual information based on the user input.
• Level 3: Geospatial Intelligence Products. Following early-warnings generated in Level 1 or 2, geospatial intelligence products requiring human intervention are provided upon user request.
The objective of ARCOS is to design and implement an early-warning system providing continuous monitoring of the Arctic Region. Designed to generate actionable products in the security domain by processing and fusing multi-sensor data, the system integrates available information from space, non-space sources and products available from multiple Copernicus services.
ARCOS provides information services at three different levels of scale and user interaction:
• Level 1: Automatic Early-warning System. Integration of space and non-space data sources for the triggering of alarms on the region when certain conditions happens. Automatic early-warnings are generated in case anomalous behaviors are detected. For this wide-area monitoring, automatic extraction of analytics and AI techniques are applied.
• Level 2: User-Driven Alert System, where space and non-space data are processed on specific locations provided by the user. The alarms can be configured as contextual information based on the user input.
• Level 3: Geospatial Intelligence Products. Following early-warnings generated in Level 1 or 2, geospatial intelligence products requiring human intervention are provided upon user request.
The project is organized in various technical activities. Currently the project at the end has reached the following main achievements:
1) Definition of Use Cases based on the analysis of user needs (extracted from Questionnaire and meetings with the Users): the use cases identify concrete scenarios on sensitive areas of Arctic region. In the Use Cases, a methodology is also defined to demonstrate the concepts of continuous monitoring (Level 1) and of service proactively producing information in the form of early warnings and alarms for the users (Level 2);
2) Experimentation and analysis of available open datasets (AIS, Sentinel 1, Sentinel 2, Sentinel 5p, CMEMS) and of state-of-the-art technology in order to design and develop the ARCOS Services;
• Capability for large monitoring of Arctic areas (Land and Sea) searching for pollution sources in order to find the possible source (e.g. a new industrial facility or gas extraction site) using Sentinel 5P;
• Capability for large monitoring of Arctic sea areas using AIS data and extracting pattern of life maps, perform anomaly analysis and extract time-series over specific AOIs;
• Capability of vessel detection specifically tailored on Arctic Region using EO data that allows the identification of dark vessels not providing AIS data (non-cooperative);
• Capability to make land-change analysis using SAR data and in particular: Sentinel 1, SAR VHR data.
3) Some of the achievements from WP3 have been integrated into the first version of the platform by performing a full production of one of the area of interests identified in the use cases. In this test area, a big amount of data have been processed from various sources (AIS, Sentinel 1) in order to provide a first version capable of showing some of the capabilities described in the Use Cases.
4) In particular, the platform development (WP4) has concentrated his efforts on the design of the platform by selecting the main web standards (OGC, STAC, OpenEO) and by defining the core architecture based on open source technology and a distributed set of nodes.
5) Development and integration of the ARCOS platform (final version v3)
6) Setup of the Gtilab repository to track project incremental development and to manage the developed code and future maintenance
7) Generation of products as output of T4.3 including final draft of processing chains descriptions including the characteristics of input/output and parameters:
o Sea Activity Anomaly Detection service
o Vessel detection services
o Oilspill detection services
o Land Change Detection and Anomalies service
o Atmospheric Pollution Monitoring service
o Arctic monitoring products
The achievements of the project have been also presented and discussed in 5 meetings with the Advisory Board in order to receive feedbacks and keep the group aligned with the project progresses and get early feedback.
1) Definition of Use Cases based on the analysis of user needs (extracted from Questionnaire and meetings with the Users): the use cases identify concrete scenarios on sensitive areas of Arctic region. In the Use Cases, a methodology is also defined to demonstrate the concepts of continuous monitoring (Level 1) and of service proactively producing information in the form of early warnings and alarms for the users (Level 2);
2) Experimentation and analysis of available open datasets (AIS, Sentinel 1, Sentinel 2, Sentinel 5p, CMEMS) and of state-of-the-art technology in order to design and develop the ARCOS Services;
• Capability for large monitoring of Arctic areas (Land and Sea) searching for pollution sources in order to find the possible source (e.g. a new industrial facility or gas extraction site) using Sentinel 5P;
• Capability for large monitoring of Arctic sea areas using AIS data and extracting pattern of life maps, perform anomaly analysis and extract time-series over specific AOIs;
• Capability of vessel detection specifically tailored on Arctic Region using EO data that allows the identification of dark vessels not providing AIS data (non-cooperative);
• Capability to make land-change analysis using SAR data and in particular: Sentinel 1, SAR VHR data.
3) Some of the achievements from WP3 have been integrated into the first version of the platform by performing a full production of one of the area of interests identified in the use cases. In this test area, a big amount of data have been processed from various sources (AIS, Sentinel 1) in order to provide a first version capable of showing some of the capabilities described in the Use Cases.
4) In particular, the platform development (WP4) has concentrated his efforts on the design of the platform by selecting the main web standards (OGC, STAC, OpenEO) and by defining the core architecture based on open source technology and a distributed set of nodes.
5) Development and integration of the ARCOS platform (final version v3)
6) Setup of the Gtilab repository to track project incremental development and to manage the developed code and future maintenance
7) Generation of products as output of T4.3 including final draft of processing chains descriptions including the characteristics of input/output and parameters:
o Sea Activity Anomaly Detection service
o Vessel detection services
o Oilspill detection services
o Land Change Detection and Anomalies service
o Atmospheric Pollution Monitoring service
o Arctic monitoring products
The achievements of the project have been also presented and discussed in 5 meetings with the Advisory Board in order to receive feedbacks and keep the group aligned with the project progresses and get early feedback.
The extreme latitude of the Arctic generates some specific challenges that are exclusive to this area of the world. Although they are all technically solvable, a comprehensive solution must take them all into consideration. Some of the challenges are:
• The very large seasonal difference in light conditions that affects optical data acquisition which is basically possible only for half of the year
• The presence of false alarms in the vessel detection due to the presence of sea-ice and icebergs in motion, and
• The default geographic information in datum and projection that ignores deformations in extreme latitudes.
The iceberg detection, and its discrimination from ships by SAR, is first of all crucial for navigation safety and, at the second time, for the identification of “dark vessels” that can be spotted with combination of AIS (e.g. Air Transmission gap detection) and Sentinel 1.
In order to overcome the necessity to analyse huge areas, AIS data can be used to detect anomalous behaviour of vessels and provide useful insights both on the vessels by detecting: Anchoring, a vessel that is stationary for a long period in open sea; Rendezvous, two vessels meeting in open sea for a long period
Also in this case, the Arctic region provides challenges like the presence of the icepack during winter months that can bring the vessel to follow a very narrow route thus resulting in false alarms in rendezvous detection. For this reason, the analysis allow to filter the events depending on the attributes specific of the event (e.g. type of vessels involved, speed and direction).
In the case of land change, the difficult conditions of Arctic (illumination, seasonality, cloud presence) gives huge challenges for the use of optical data. SAR sensors have been thus be experimented to detect possible anomalies and in general for all monitoring purposes. In general, the progress will be based on:
• Detection of changes using both HR images (Sentinel 1) and VHR images (Cosmo-Skymed, Terrasar, Iceye) and in particular MTC, ACD and CCD maps in a continuous way
• Detection of routes in the ice-pack and of populated areas (to be further investigated)
• Automatic change detection in a specific Area of Interest (to be further investigated) using EO data and available open datasets like OpenStreetMap
• Detection of Atmospheric pollution anomalies to provide alerting
• An analysis of Social Media and News sources through language analysis and filtering (to be further investigated).
• The very large seasonal difference in light conditions that affects optical data acquisition which is basically possible only for half of the year
• The presence of false alarms in the vessel detection due to the presence of sea-ice and icebergs in motion, and
• The default geographic information in datum and projection that ignores deformations in extreme latitudes.
The iceberg detection, and its discrimination from ships by SAR, is first of all crucial for navigation safety and, at the second time, for the identification of “dark vessels” that can be spotted with combination of AIS (e.g. Air Transmission gap detection) and Sentinel 1.
In order to overcome the necessity to analyse huge areas, AIS data can be used to detect anomalous behaviour of vessels and provide useful insights both on the vessels by detecting: Anchoring, a vessel that is stationary for a long period in open sea; Rendezvous, two vessels meeting in open sea for a long period
Also in this case, the Arctic region provides challenges like the presence of the icepack during winter months that can bring the vessel to follow a very narrow route thus resulting in false alarms in rendezvous detection. For this reason, the analysis allow to filter the events depending on the attributes specific of the event (e.g. type of vessels involved, speed and direction).
In the case of land change, the difficult conditions of Arctic (illumination, seasonality, cloud presence) gives huge challenges for the use of optical data. SAR sensors have been thus be experimented to detect possible anomalies and in general for all monitoring purposes. In general, the progress will be based on:
• Detection of changes using both HR images (Sentinel 1) and VHR images (Cosmo-Skymed, Terrasar, Iceye) and in particular MTC, ACD and CCD maps in a continuous way
• Detection of routes in the ice-pack and of populated areas (to be further investigated)
• Automatic change detection in a specific Area of Interest (to be further investigated) using EO data and available open datasets like OpenStreetMap
• Detection of Atmospheric pollution anomalies to provide alerting
• An analysis of Social Media and News sources through language analysis and filtering (to be further investigated).