ABSTRACT
The rapid development of remote sensing technology has increased the possibility to produce a spatially explicit estimation of Total Suspended Solids (TSS). However, factual information regarding the effective depth of waters where TSS can be effectively mapped using remote sensing data is rare. This information is essential to determine the effective depth of TSS mapping and the effect of the water column on the representativeness of the estimation results. This study aimed to (1) determine the effective water depths of TSS mapping using PlanetScope imagery, (2) map the spatial distribution of TSS at every effective depth using PlanetScope imagery, and (3) analyse the spatial distribution of the mapped TSS based on the best empirical modelling for every effective depth. PlanetScope image offered new opportunities in high spatial resolution remote sensing with daily revisit time. Four single bands of PlanetScope images (blue, green, red, and near-infrared) – corrected to the Bottom of Atmosphere (BOA) reflectance, 12 band ratios, and four Principal Component bands (PC-band) were inputted to the determination process of the effective water depths. Empirical modelling between the in-situ TSS at each effective water depth andures what was the issue, wasnt able to o PlanetScope pixels at the corresponding locations was conducted. However, only image bands that exceeded the significance limit (r) were used for modelling using linear, exponential, logarithmic, second-order polynomial, or power regressions. The results showed that the band ratios of PlanetScope images could record TSS up to the effective depth of 1.8 m. The best empirical modelling in each effective depth of waters was the band ratio which is dominated by the contribution of the blue bands (B1), red (B3), and NIR (B4). B4/B3 bands combination produced TSS information at the effective depths of 0–0.2 and 0–0.4 m, B3/B4 bands at 0–0.6 and 0–0.8 m, B1/B4 bands at 0–1 and 0–1.2 m, and B1/B3 bands at 0–1.4, 0–1.61, and 0–1.8 m. For all effective depths, the spatial distribution pattern of TSS in the study site (Menjer Lake, Central Java, Indonesia) showed that high concentrations evenly spread along the edge and became increasingly lower to the centre of the lake. Meanwhile, the vertical distribution showed that the deeper the cumulative water depth, the higher the TSS.
Acknowledgements
The authors would like to thank (1) Master of Science Programme in Remote Sensing at the Faculty of Geography, Universitas Gadjah Mada, Indonesia, for providing research facilities, (2) PNBP Project from Indonesia Power-Mrica Unit, and Limnology Research Center (P2L) of the Indonesian Research Institute (LIPI) for assisting in field activities, and (3) Planet Labs Inc. for providing access to PlanetScope images. This work was supported by Rekognisi Tugas Akhir (RTA – final student project recognition) 2020 from Universitas Gadjah Mada, Indonesia, under Grant number 2488/UN1.P.III/DIT-LIT/PT/2020.
Data Availability Statement
The research data for this paper can be accessed through the first author ([email protected]).
Disclosure Statement
No potential conflict of interest was reported by the author(s).