- TPM data help advance plastic ...
TPM data help advance plastic litter detection from space
05 Jun 2023
Several innovative ESA Discovery element projects exploring remote sensing of plastic marine litter lean on commercial satellite data disseminated via the Agency’s Third Party Mission programme.
Plastic litter impacts marine ecosystems, human health and maritime safety. Detecting and tracking marine plastic debris using Earth observation (EO) data can help monitor this environmental problem and its mitigation but remote sensing technology and research in the area are in their early stages.
The data can help pin-point major sources of litter, assess in a timely fashion the effectiveness of mitigation policies and interventions to limit waste leaking from into sea, and possibly optimise clean-up strategies.
In the hope of accelerating international research to tackle the issue of marine plastic, ESA launched a call for ideas in 2019 – the first of its kind – for monitoring marine litter from space.
As a result, 26 research projects were initiated by the Discovery element of ESA’s Basic Activities, with novel ideas to detect and possibly measure plastic concentrations and help identify the distribution and sources of marine plastic pollution at a global level.
Many of these projects relied on commercial very high-resolution imagery from non-ESA missions, such as WorldView, PlanetScope, ICEYE and Pléiades, to boost their activities. Data from these Third Party Missions (TPM) are available for free for research and application development purposes via ESA’s TPM programme.
One challenge with monitoring floating plastic litter is that it is continuously moving. The Discovery element’s TRACE project aimed to overcome this issue by using both high-resolution optical data from PlanetScope and Synthetic Aperture Radar (SAR) data from Copernicus Sentinel-1, coupled with transport computer modelling.
The research team, led by GFZ Potsdam, Germany, successfully developed a prototype for a real-time detection and tracking system, which could obtain precise and reliable information on the position and movement of floating macro-litter and litter accumulation patches.
As the Mediterranean Sea is a well-known global hotspot for marine plastic litter, the study focused on the northern part of the Adriatic. They used a time series of PlanetScope and Sentinel-1 data for the period from 28 May to 3 June 2021.
The team demonstrated that coupling daily remote sensing data with oceanic forecasting models, enables the distinction between geo-stable objects (e.g., buoys and small rocks), actively driving objects (e.g., boats) and passively drifting objects, such as marine litter. It is hoped the prototype will eventually lead to a fully operative marine litter monitoring system.
Invasive aquatic weeds help track riverine plastic
Although rivers function as pathways to transport plastic from the land to the sea, the study of riverine plastic via remote sensing remains in its infancy. The ‘Plastic Monitor’ project, led by Deltares, the Netherlands, aimed at understanding the feasibility of detecting riverine plastic accumulations in Indonesia – a country where plastic pollution is a pressing concern.
As well as characterising floating plastic accumulation patches in the Indonesian Citarum River, the study compared satellite imagery of known floating debris accumulation patches with in-situ camera images. A synergy of EO data from Copernicus Sentinel-1, Sentinel-2 and TPM satellites – WorldView-3 and ICEYE – were used. Water hyacinths play an important role in gathering and transporting macro plastic litter in riverine ecosystems, as these fast-growing and invasive freshwater plants tend to form large patches at the water surface.
A research team led by the University of Wageningen, the Netherlands, showed that floating river plastics are detectable within hyacinth patches in the Saigon River, around Ho Chi Minh City, Vietnam, using high spatial resolution Worldview-3 multispectral optical panchromatic imagery. In the Saigon River, plastic debris were successfully detected within hyacinth patches using Maxar’s Worldview-3 multispectral optical and panchromatic imagery (with a ground sample distance of ∼1.2 m and ∼0.3 m respectively).
AI and drones support mapping of marine plastic
Another Discovery element project studied the feasibility of a technical solution based on artificial intelligence and drones, where results are compared to processed images from very high-resolution TPM data. For remote sensing detection of marine plastic litter, sufficiently high revisit times and spatial resolutions are required.
Consequently, the AIDMAP (Artificial Intelligence and drones supporting the detection and mapping of floating aquatic plastic litter) project led by VITO, Belgium, proposed an approach combining information from multispectral cameras on small drones, satellites (such as WorldView, Pléiades and PlanetScope) and in-situ observations, with artificial intelligence (AI), to detect floating and inland plastic litter.
Plastic’s characteristic spectral reflectance can help distinguish it from other materials, such as wood, vegetation and surface features. However, discrimination can become difficult when the plastic is slightly submerged, wet, or otherwise mixed with different materials and in relatively low concentration within the single image pixel.
In the FRONTAL project, led by the Plymouth Marine Laboratory, UK, very high-resolution imagery from Worldview-3 was used to validate Copernicus Sentinel-2 detections of accumulations of floating marine plastic debris on ocean fronts, which appear at the boundary between water masses with different properties. This call for ideas and subsequent activities were led by the Discovery element of ESA’s Basic Activities through the Open Space Innovation Platform (OSIP), ESA's hub for sourcing ideas from academia, industry and the general public.
TPM data are disseminated via ESA’s TPM programme through agreements with international public and commercial data owners.
Learn about ESA's TPM data and access the collections.