The use of 'optical water type' classification schemes is becoming increasingly prevalent within limnological and oceanographic remote sensing research (Moore et al. 2001, 2014, Jackson et al. 2017, Spyrakos et al. 2018). Uses for optical water classes include, but are not limited to, algorithm blending (Moore et al. 2001), product uncertainty estimation (Jackson et al. 2017), data quality flagging (Wei et al. 2016) water quality monitoring (Uudeberg et al. 2020) and environmental phenology (Trochta et al. 2015) studies.
However, a harmonised approach to the creation and use of the classes has not yet emerged from the research community. Despite recent efforts to move to a unified fuzzy logic scheme (Jai et al. 2021), a diversity of distance metrics, data transformations and cluster optimisation schemes are applied at local scales (Bi et al. 2019, Botha et al. 2020, da Silva et al. 2020, Uudeberg et al. 2020). Though all these approaches provide interesting and useful results, the fragmented nature of the research makes the comparison of water types difficult, impeding collaboration and optimisation of methods.
As with most machine learning techniques, unsupervised clustering is susceptible to the problems of insufficient or biased training data, the ‘central tendency’ (Malik, 2020), and overtraining. Here we examine components of the clustering pipeline such as data transformation, data dimensionality, distance metric choice and output cluster variables to present a generalised approach that uses robust, data-driven principles to generate a cluster set with minimal 'human' decision making. The approach presented builds upon insight from two European Space Agency projects (Ocean Colour Climate Change Initiative, OC-CCI; Lakes-CCI). The approach is demonstrated using Sentinel 2 MSI and Sentinel 3 OLCI data, at the regional and pan-regional scale, across a range of optical environments. Matchups against in-situ measurements are used to validate the utility of the clusters generated.
This work was undertaken as part of the EC Horizon 2020 CERTO (Copernicus Evolution: Research for harmonized Transitional water Observation) project.
Bi, S, Li, Y., Xu, J., Liu, G., Song, K., Mu, M., Lyu, H., Miao, S., and Xu, J. "Optical classification of inland waters based on an improved Fuzzy C-Means method," Opt. Express 27, 34838-34856 (2019)
Botha, Elizabeth & Anstee, Janet & Sagar, Stephen & Lehmann, Eric & Galvao, Thais. (2020). Classification of Australian Waterbodies across a Wide Range of Optical Water Types. Remote Sensing. 12. 3018. 10.3390/rs12183018.
Jackson, T., Sathyendranath, S. and Mélin, F. “An improved optical classification scheme for the Ocean Colour Essential Climate Variable and its applications,” Remote Sens. Environ. 203, 152–161 (2017).
Jia T, Zhang Y, Dong R. A Universal Fuzzy Logic Optical Water Type Scheme for the Global Oceans. Remote Sensing. 2021; 13(19):4018. https://doi.org/10.3390/rs13194018
Malik, Momin M. "A hierarchy of limitations in machine learning." arXiv preprint arXiv:2002.05193 (2020).
Moore, Timothy & Campbell, Janet & Feng, Hui. (2001). A fuzzy logic classification scheme for selecting and blending satellite ocean color algorithms. IEEE Transactions on Geoscience and Remote Sensing, 39, 1764-1776. Geoscience and Remote Sensing, IEEE Transactions on. 39. 1764 - 1776. 10.1109/36.942555.
Moore, Timothy & Dowell, Mark & Bradt, Shane & Ruiz-Verdu, Antonio. (2014). An optical water type framework for selecting and blending retrievals from bio-optical algorithms in lakes and coastal waters. Remote sensing of environment. 143. 97-111. 10.1016/j.rse.2013.11.021.
Spyrakos, E., O’Donnell, R. , Hunter, P. D. , Miller, C. , Scott, M., Simis, S. G., Neil, C., Barbosa, C. C., Binding, C. E. and Bradt, S. “Optical types of inland and coastal waters,” Limnol. Oceanogr. 63(2), 846–870 (2018).
da Silva, E.F.F.; Novo, E.M.L.d.M.; Lobo, F.; Barbosa, C.; Noernberg, M.A.; Rotta, L.H.d.S.; Cairo, C.T.; Maciel, D.A.; Flores Júnior,R. Optical water types found in Brazilian waters. Limnology (2020).
Trochta, John & Mouw, Colleen & Moore, Timothy. (2015). Remote sensing of physical cycles in Lake Superior using a spatio-temporal analysis of optical water typologies. Remote Sensing of Environment. 171. 10.1016/j.rse.2015.10.008.
Uudeberg, K.; Aavaste, A.; Kõks, K.-L.; Ansper, A.; Uusõue, M.; Kangro, K.; Ansko, I.; Ligi, M.; Toming, K.; Reinart, A. Optical Water Type Guided Approach to Estimate Optical Water Quality Parameters. Remote Sens. 2020, 12, 931. https://doi.org/10.3390/rs12060931
Wei, Jianwei & Lee, Zhongping & Shang, Shaoling. (2016). A system to measure the data quality of spectral remote sensing reflectance of aquatic environments. Journal of Geophysical Research: Oceans. 10.1002/2016JC012126.