Direct detection of submerged plastics using hyperspectral imaging technology
–Advancement to eliminate water interference–

1. Key Points

• The hyperspectral imaging technique^※1 using near-infrared light is effective for identifying plastic materials and is a useful technology for evaluating marine plastics. However, to date, water interference effects have complicated identifying these materials, owing to the loss of light reflection properties.

• In this study, plastic materials were successfully identified through analysis using the principle of differential optical absorption spectroscopy^※2, even in environments where water is present. Nine types of materials, which account for the composition of 72% of the world’s plastic waste, could be identified when submerged in water to a depth of approximately 10 mm.

• The results of this study demonstrate the potential for detecting marine plastics in real submerged environments and can serve as fundamental knowledge for direct observation using drones, aircraft, and satellites. It is expected that these results will be applied as a basis for developing new approaches to elucidate the full extent of marine plastic pollution.

2. Overview

Researcher Chunmao Zhu of the Earth Surface System Research Center (ESS), Research Institute for Global Change (RIGC), Japan Agency for Marine-Earth Science and Technology (JAMSTEC), along with Director and Principal Researcher Yugo Kanaya, developed a method for detecting submerged plastics by combining differential optical absorption spectroscopy with a hyperspectral imaging technique. This method is a key element needed for directly detecting marine plastics by removing water interference.

The severity of the negative effects of marine plastic waste is unquestionable in modern society, while mass consumption of plastic is increasing. Efficient detection methods are needed to fully understand the current distribution of marine plastic waste. The efficacy of hyperspectral measurements has previously been demonstrated, but due to water absorbing the near-infrared light that is used for hyperspectral imaging diagnosis, its applicability has been limited to dried plastic samples after filtration.

In this study, the detectivity of wet plastic samples present on the water surface, or those present beneath the water surface, was evaluated by obtaining reflectance spectra of plastic pieces submerged in water from 2.5 mm to 15 mm deep, in a laboratory experiment. A method was then developed to remove the effects of water in the analysis spectra, and its effectiveness was confirmed. Furthermore, the optimal wavelength range for the detection of nine types of polymers that account for the composition of 72% of the world’s plastic waste was identified.

The results of this study suggest a new approach that is effective in directly detecting submerged plastics, which has been difficult to date. This is an important step in establishing an innovative measurement technique to understand the current state of marine plastic pollution, and new prospects can be expected for evaluating the impact on the marine environment.

This study was conducted under the Ministry of Education, Culture, Sports, Science, and Technology commissioned project “Program for the Development of Technology to Promote the Utilization of Ocean Resources” Research ＆ Development Proposal for Determining Marine Information “Development of automatic microplastic analysis method using a hyperspectral camera” (Grant number: JPMXD0618067484).

These results are scheduled to be published in “Scientific Reports” on October 6 (Japan time).