Reproducible pipelines and readiness levels in plastic monitoring

Flexible decision-making tools are needed to support action plans for plastics and other pollutants. Reproducible Analytical Pipelines (RAPs) and technological readiness levels (TRLs) will enable systematic validation and global harmonization of plastic pollution monitoring methods.

Flexible decision-making tools are needed to support action plans for plastics and other pollutants. Reproducible Analytical Pipelines (RAPs) and technological readiness levels (TRLs) will enable systematic validation and global harmonization of plastic pollution monitoring methods.
Plastic pollution is a wicked problem 1 that spans all environmental compartments, with different magnitudes in space and time. A Global Plastic Treaty is under preparation 2 with the ambitious goal of producing a set of legally binding tools aimed at stopping or reducing the flow of plastics into the environment. Policymakers and scientists are looking forward to endorsing monitoring plans based upon ready-to-deploy methods for different analytical scenarios. However, plastic monitoring is facing a reproducibility crisis 3 . Despite attempts to define monitoring guidelines, there are still no widely accepted monitoring frameworks. Tools and protocols have been developed to quantify plastic pollution, but these methods often provide incomparable results, even if applied to the same environmental matrix 4 .
To promote and accelerate the adoption of best monitoring practices, a flexible method-validation framework based on reproducibility, replicability, and repeatability 5 is urgently required. In this Comment, we propose the application of RAPs and TRLs as a tool to support policy and technical decisions about plastic monitoring.

Reproducible analytical pipelines
RAPs are a set of automated processes used to identify best practices needed to assure that coding pipelines and data processing are standardized, quality controlled and reproducible. The concept was first introduced to manage workflows in software engineering and it is now widely applied to streamline industrial processes 6 . RAPs are especially helpful for multilevel workflows (like many plastic monitoring methods), providing modularity as a possible solution 7 .
At present, each plastic monitoring guideline is traditionally considered a unique, solid and complete path dedicated to a single matrix and particle size. Moving forwards, we advocate framing these workflows as modular RAPs, where any methodological step is separately evaluated and then implemented, saving money and time compared with evaluating a full pipeline.
Plastic monitoring can be divided into six modules in the RAP: survey design, sample collection, sample preparation, analytical detection, quantification, and data reporting (Fig. 1a). Important information can be extracted when every step in a RAP is investigated separately. For instance, scientists or policymakers can decide if a single step in the RAP (such as the use of analytical instruments to confirm the polymeric identity of particles) is mature enough to be implemented in all monitoring guidelines that share it. If the method is not mature, further testing and validation can be recommended.
To support this decision-making, it is important to use a robust and synthetic approach to assess the maturity of each step of a plastic monitoring RAP (that is, how much a technology is ready to fulfil the expected tasks). Although rarely applied to environmental science 8 , we suggest using TRLs -developed by NASA to evaluate if a space technology was ready for deployment or needed further development 9 -for this assessment.

Technological readiness level
The TRL scale classifies technology or methods into basic research (TRLs 1-3), applied research (TRLs 4-5), in development (TRLs 6-8) and implementation (TRL 9) phases (Fig. 1b). Where a technology falls on the scale is usually assessed by experts' opinions. In plastic research and monitoring, TRL can be based on the functionality, reliability, usability, efficiency, maintainability, accessibility, cost, and portability of a method. These aspects could be ranked and assessed using a SWOT (strengths, weaknesses, opportunities, and threats) approach. The outputs of these systematic assessments should be freely available to relevant stakeholders, deposited in suitable open-access repositories (such as the GPML digital platform, and repeated and updated on a regular basis. This information will support informed decision-making, but before implementation, scientific, technical, logistical, environmental, and ethical constraints must be considered 10 .

Merging RAPs and TRLs
The TRL approach could be simply applied to entire full plastics monitoring guidelines; however, we argue that if applied singularly to each step in a RAP, it has the potential to greatly improve and accelerate the selection, evaluation, and adoption of large-scale plastic monitoring programmes.
For instance, no methodological standards exist for microplastic sampling in the air (for example, using active versus passive samplers, measuring dry versus wet deposition, and appropriate sampling volume and duration). Therefore, air sampling-related modules would have a TRL <3, as they are still at a basic research level and not yet ready for monitoring recommendation. Conversely, analysis of samples with Fourier transform infrared (FTIR) spectroscopy is not dependent on the sampling method or matrix, and is commonly used for plastic polymer identification. FTIR would have a TRL of 9 and could be recommended for air monitoring guidelines. Overall, the low TRL of the sampling module prevents the definition of a full standard pipeline for monitoring microplastics in the air, but breaking the method down into the