Inherent optical properties of suspended particulate matter in the southern Baltic Sea*

The inherent optical properties (IOPs) of suspended particulate matter and their relations with the main biogeochemical characteristics of particles have been examined in the surface waters of the southern Baltic Sea. The empirical data were gathered at over 300 stations in open Baltic Sea waters as well as in the coastal waters of the Gulf of Gdańsk. The measurements included IOPs such as the absorption coefficient of particles, absorption coefficient of phytoplankton, scattering and backscattering coefficients of particles, as well as biogeochemical characteristics of suspended matter such as concentrations of suspended particulate matter (SPM), particulate organic matter (POM), particulate organic carbon (POC) and chlorophyll a (Chl a). Our data documented the very extensive variability in the study area of particle concentration measures and IOPs (up to two orders of magnitude). Although most of the particle populations encountered were composed primarily of organic matter (av. POM/SPM = ca 0.8), the different particle concentration ratios suggest that the particle composition varied significantly. The relations between the optical properties and biogeochemical parameters of suspended matter were examined. We found significant variability in the constituent-specific IOPs (coefficients of variation (CVs) of at least 30% to 40%, usually more than 50%). Simple best-fit relations between any given IOP versus any constituent concentration parameter also highlighted the significant statistical errors involved. As a result, we conclude that for southern Baltic samples an easy yet precise quantification of particle IOPs in terms of the concentration of only one of the following parameters – SPM, POM, POC or Chl a – is not achievable. Nevertheless, we present a set of best statistical formulas for a rough estimate of certain seawater constituent concentrations based on relatively easily measurable values of seawater IOPs. These equations can be implemented in practice, but their application will inevitably entail effective statistical errors of estimation of the order of 50% or more.