An Experiment on Satellite Remote Sensing of Plastic Waste in Pasig River

By Rex Lor, Head of Solutions Mapping, Accelerator Lab Philippines

September 29, 2021

 

It’s the typhoon season again. Plastics, which had lain dormant during calmer weather, have now come back to haunt us.  

According to a report, Metro Manila’s largest pumping station has been damaged by an estimated 2 tons of trash being collected daily. Majority of these are plastic waste washed into Pasig river by torrential rainfall.

 

MMDA Officials inspect solid waste collected from a pumping station in Metro Manila. (Source: www.inquirer.net)

The huge solid waste problem in the Philippines is highlighted by a research conducted by Jambeck et al. in 2015 where the Philippines was ranked as one of the top contributors of marine litter in Asia, throwing 2.7 million tons of plastic into the sea every year.  Pasig River, dubbed as one of world’s top plastic polluters, accounts for 21% of the organic waste flow to Manila Bay, 70% of which come from households. The river is a major waterway that flows from Laguna Lake in the east to Manila Bay in the west, via five independent city administrations in the central part of Metro Manila (Pasig, Mandaluyong, Taguig, Makati, Manila). 

Considering the amount of rubbish that ends up in the pumping stations, it should come as no surprise that they are damaged during the monsoon season.  The stations bear the burden of the enormous amounts of solid waste that are dumped in their path.  Further complicating this problem is the increase of single-use plastics and disposable personal protective equipment (PPE) generating an estimated 52,000 metric tons of medical waste amid the pandemic.

With the complex problem of solid waste and plastic pollution in Metro Manila layered with the complications of the COVID-19 pandemic, robust data is needed to better understand waste flow through cross-boundary tributaries that ultimately lead to Manila Bay. However, traditional field work is challenging at this time, with quarantine restrictions and risks to COVID-19 exposure inhibiting physical movement. This complex challenge led us to ask an important this question: Can existing satellite imagery data provide us with a safe, efficient and accurate method of generating insights on plastic waste production in Metro Manila? 

This desire came into fruition with the Japan SDGs Innovation Challenge where we were given the opportunity to conduct this type of experiment.  Supported by the Cabinet Office of the Government of Japan, the UNDP PH Accelerated Lab (ALabPH) engaged the Japan Manned Space Systems Corporation (JAMSS) to build a remote sensing model to detect plastic litter in the Metro Manila river system using available satellite image data.  The project was also assisted by the Japan Innovation Network (JIN)

JAMSS used the spectral unmixing method that solves the spectral response as an inverse problem and can estimate the abundance of the target substance from the spectrum.

To test the spectral image sensing model developed by JAMSS, we identified 10 areas in the Metro Manila River system.  The areas include: 1) Estero de Binondo, Manila; 2) Sevilla Bridge, San Juan; 3) Estero de Galina, Pasay; 4) Estero de Maypajo, Navotas; 5) Estero de Vitas, Manila; 6) Ilugin River intersecting Parian Creek, Pasig; 7) Marikina-Pasig River intersection, Pasig; 8) Maybunga Floodway, Pasig; 9) Pasig River near Guadalupe Bridge, Makati; 10) Marikina River, Marikina-Pasig.

As we were working closely with the Pasig City Environment and Natural Resources Office (Pasig CENRO), 5 of the 10 areas focused within the boundary of the city.  The area around Pasig City is also an interesting location as it is a confluence of different canals and rivers making it a natural catch basin.  This also allows us to view the state of the canal and river system around Pasig city from the vantage point 800 kilometers above. 

Key Findings and Lessons Learned

As the first phase focuses more on the development of the image sensing model, here are some of the lessons we learned from the implementation of the project:

 
  • A majority of the ten initial areas yielded mostly positive results based on the JAMSS report and the spectral image sensing model was able to identify and locate plastics.  It is worth noting that specific areas in Pasig City yielded ‘uncertain’ results during the satellite remote sensing activity and the ensuing validation activities conducted by the joint Pasig City and UNDP PH drone team. Further inquiry into the matter with the Pasig CENRO indicated that the regular river and estuary clean up conducted by the city has yielded positive results as the satellite remote sensing was no longer able to detect plastic waste. 

 
  • The remote sensing model is as good as the available resolution of the satellite images.  Because we are constrained with the available satellite image resolution, coverage of canals and estuaries is inadequate and can only get the desired accuracy up to 5-10 square meters.  In addition, most canals and estuaries are covered by either foliage (e.g. trees, bushes, and river hyacinth) or roofs of illegal settlers.
  • Which brings us to the direction of the second phase to focus more on the larger river systems in Metro Manila that have at least 5 square meters. In other words, we need to scale out the coverage to cover other river systems such as, but not limited to, Pasig River, Tullahan River, Marikina River, San Juan River, Paranaque River, Manggahan Floodway, Tunasan River, Taguig River, among others.
  • To cover smaller canals and estuaries, the planned Phase 2 includes conducting an experiment on using citizen science to augment information provided by the satellite remote sensing project.  This will also provide baseline information on marine plastic pollution in Metro Manila and raise awareness on the ill effects of marine plastic pollution by engaging the citizen researchers.
  • As phase 2 looks to scale up and scale out the solution, there is a need to engage and involve more key stakeholders, especially from national government agencies, research institutions, and Metro Manila cities.  Proper guidance and direction can be gained especially from the government partners as they may have conducted similar research using satellite remote sensing technology.

Moving Forward

The first phase of the satellite remote sensing experiment provided us with essential lessons that will guide us in the phase 2 strategy of scaling out, scaling up, and scaling deep.  By scaling out, we plan to focus on larger river systems in Metro Manila including Manila Bay.  By scaling up, we plan to explore other methodologies to layer the spectral information image sensing model already developed by JAMSS.  By scaling deep, we plan to reach out to key government agencies and research institutions to learn from their experiences and enhance the experiment.   

We hope that by pursuing this project, we would be able to provide a way to monitor our rivers over time to better inform inter-city regulation, policy harmonisation, and align implementation of monitoring and clean-up drives.  

The use of earth observation satellites for remote sensing bears significant potential to become a great tool to guide development policymaking. In which other ways do you think we can best use this cutting-edge innovation? We will be happy to hear from you! E-mail us at acceleratorlab.ph@gmail.com