EGU21-12085
https://doi.org/10.5194/egusphere-egu21-12085
EGU General Assembly 2021
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Monitoring plastic accumulation in water hyacinths using remote sensing

Niels Janssens1, Lauren Biermann2, Louise Schreyers3, Martin Herold1, and Tim van Emmerik3
Niels Janssens et al.
  • 1Laboratory of Geo-information Science and Remote Sensing, Wageningen University
  • 2Plymouth marine laboratory
  • 3Hydrology and quantitative water management group, Wageningen University

While efforts to quantify plastic waste accumulation in the marine environment are rapidly increasing, the data on plastic transport in rivers are relatively scarce. Rivers are a major source of plastic waste into the oceans and understanding seasonal dynamics of macroplastic transport is necessary to develop effective mitigation measures. Macroplastic transport in rivers varies significantly throughout the year. Research shows that in the case of the Saigon river, Vietnam, these plastic transport fluxes are mainly correlated to the amount of organic debris (mostly water hyacinths). Since large water hyacinths patches can be monitored from space, this gives the opportunity for large scale monitoring using freely available remote sensing products. Remote sensing products, such as Sentinel-2, can be applied to areas where water hyacinths occur and plastic emissions are estimated to be high. In this study, we present a first method to detect and monitor water hyacinths using optical remote sensing. This was done by developing an algorithm to automatically detect and quantify water hyacinth coverage for a large section of the Saigon river in Vietnam, for the year 2018. Spectral signatures of water,  infrastructure in the river, and water hyacinths were used to classify the water hyacinths coverage and dynamics using a Naive Bayes algorithm. Water hyacinths were promisingly identified with 95% accuracy by the Naive Bayes classifier. The comparison between the seasonal dynamics of classified water hyacinth and seasonal dynamics of the field measurements resulted in an overall Pearson correlation of 0.72. The comparison we attempted between seasonal dynamics of plastics from satellite and field measurements yielded a Pearson correlation of 0.48. With the next field campaign collecting in-situ data matched to satellite overpasses, we aim to improve this. In conclusion, we were able to successfully map seasonal dynamics of water hyacinth in an automated way using Sentinel-2 data. Our study provides the first step in exploring the possibilities of mapping water hyacinth from satellite as a proxy for river plastics.

How to cite: Janssens, N., Biermann, L., Schreyers, L., Herold, M., and van Emmerik, T.: Monitoring plastic accumulation in water hyacinths using remote sensing, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12085, https://doi.org/10.5194/egusphere-egu21-12085, 2021.