The effect of the specific inherent optical properties on assessing water quality in Dutch inland waters using matrix inversion

Optical remote sensing is increasingly being used to detect and monitor optical water quality aspects of lakes, rivers and coastal waters. The optical models that are used for water quality retrieval need specific inherent optical properties (SIOPs) as input. However, these SIOPs may be variable between water type. In this paper, SIOPs averaged for Dutch rivers and shallow lakes were selected. Using a relatively simple optical model the effect of the variation in SIOPs on the modelled subsurface reflectance is investigated for a clear and turbid water case. Significant differences were seen between rivers and shallow lakes. Investigation of the inversion of the used optical model showed that errors in the retrieved concentrations introduced due to variability in the SIOPs can be up to 50%. This demonstrates the need for the accurate determination of the SIOPs for use in optical models, and the need for the flexibility of inversion methods to use different sets of SIOPs