Regional challenges concerning derivation of suspended particulate matter concentration and water turbidity from water reflectance. A case study in the western Black Sea

Assessment of water quality indicators, such as Suspended Particulate Matter (SPM) concentration and water turbidity (TUR) is essential for marine ecosystems health evaluations. This work focuses on the estimation of these two variables from water-leaving reflectance measurements, either derived from in-situ observations or satellite data, for the western Black Sea basin. The regional characteristics of the water optically active constituents can have important impact on the quality of such estimations if not properly accounted for. We test several existing inversion algorithms and quantify their accuracy. New, regionally adapted methods are then proposed, together with a new formulation for a multi-conditional switching mechanism. Calibration of these models is performed based on SPM and TUR in-situ measurements. It is shown that the improvements achieved through regional adaptation can be significant. Application of these algorithms to Sentinel-3 OLCI satellite data reveals the consistency of the proposed methodology. Also, it shows the degree of uncertainty if improper formulations are used, with major impact on the absolute values of retrieved SPM or TUR. •Existing algorithms for estimation of SPM and turbidity are tested using in-situ measurements, for the western Black Sea.•New parameterizations are proposed, to increase the accuracy of predicted variables based on water reflectance data.•Multi-conditional approaches are used, with adapted thresholds for switching between different wavelengths.•Proposed algorithms are applied to Sentinel-3 images and the relative differences between results are assessed.