Estimation of Suspended Particulate Matter Using Landsat 9 Imagery: Generating Algorithms and Spatio-Temporal Distributions

Abstract

The fluctuation of suspended particulate matter (SPM) is essential to the biogeochemical cycle and ecological health of coastal waters. Anthropogenic activities potentially trigger an increase in SPM, so it needs to be monitored continuously. Spatial and temporal monitoring of SPM can be carried out cost-effectively with broad coverage using a remote sensing application. This study aims to build the SPM algorithm and estimate its temporal variability. The algorithm model in this study is based on an empirical formula between field data and reflectance data with the same acquisition. Water samples were taken from 100 stations in July 2022. Half were used for model building and the other for model validation. Suspended Particulate Matter was determined gravimetrically and estimated their temporal variability was based on Landsat 9 image records from December 2021 - November 2022. The results of the analysis show that the best algorithm for SPM estimation can be built based on coastal aerosol canals (B1), blue canals (B2), and green canals (B3) with the accuracy test result of (R² = 0.68; RMSE = 5.551 mg.L^-1; MAPE= 7.07%; Bias= 0.28). The SPM temporal fluctuations were generally higher in the west monsoon and lowered in the east monsoon, ranging from 30 to 70 mg.L^-1. On the other hand, the spatial distribution shows a higher magnitude in the estuary than in the offshore waters, with a deviation of about 30 mg.L^-1. Regional authorities can use the results obtained to improve the management of coastal water quality and monitoring systems.