This pre-core indicator and its threshold values are yet to be commonly agreed in HELCOM. The indictor is included as a test indicator for the purposes of the 'State of the Baltic Sea' report 2018, and the results are to be considered as intermediate.
The indicator uses assessment unit specific threshold values that are presented as normalized values. The indicator is based on two parameters: (1) cyanobacterial surface accumulation (CSA) and (2) cyanobacterial biomass. A threshold value is set for each parameter for each assessment unit, and the combined indicator threshold value is an average of the two. Values that are above the threshold value indicate good status (Thresholds and Status evaluation figure 1).
Thresholds and Status evaluation figure 1. Schematic representation of the averaged and normalized threshold value applied in the cyanobacteria bloom index pre-core indicator, the threshold values are assessment unit specific (see Thresholds and Status evaluation table 1).
If either parameter is not applicable to a specific assessment unit, then only one parameter is used as the threshold value. If both parameters are applicable, then an averaged normalized value is used as the threshold value. The specific values for each parameter and the proposed overall threshold value are listed in Thresholds and Status evaluation 1.
Thresholds and Status evaluation table 1. Summary table of assessment unit specific threshold values in the open-sea areas.
The threshold values for the assessment units were derived separately for the two long-term datasets used for the two indicator parameters:
1 - cyanobacterial surface accumulations (CSA); the independent satellite-based time series on algae accumulations from 1979-2014 in the Baltic Sea by Kahru and Elmgren (2014) and
2 - biomass; data on in-situ observations of cyanobacteria biomass 1990-2015 collated by the HELCOM PEG group (Wasmund et al. 2015).
The threshold values are derived by combining statistical analysis of long-term data with expert judgement. The main concern in proposing threshold values is the lack of sound and consistent historical data. The data starting from 1970s or 1980s, which are available in the threshold setting datasets do not correspond to "no, or very little human impact" that could be used as a reference condition. As cyanobacterial blooms are a natural phenomenon in the Baltic Sea the threshold values do not need to describe a status with no blooms, but rather a status without extensive and potentially harmful blooms. This status should be consistent with sustainable use of the sea by humans.
Investigations of stable nitrogen and carbon isotope composition as well as of the organic carbon content of sediments have shown that the increase in nutrients and productivity began already in the 1950s–1960s (Andrén et al., 2000; Struck et al., 2000; Poutanen & Nikkilä, 2001). According to Finni et al. (2001), cyanobacterial blooms have become common in the open sea in both the Baltic Proper and the Gulf of Finland since at least the 1960s.
To meet this aim of identifying time periods with low bloom intensity, the thresholds were derived based on the time periods within the available datasets where the status was already impacted by eutrophication but the bloom intensity was low. Within the time periods covered by the available datasets, the status covers varying levels of cyanobacterial blooms. To distinguish the shorter periods of low bloom intensity in comparison to the general level, the shift detection method based on an algorithm (Rodinov 2004) was used; or if no such periods were distinguished, the averages of separate years with lower bloom intensity were calculated using the quartile method. The shift-detection method was successfully applied to all assessment units for the satellite-based cyanobacterial surface accumulations (CSA) (parameter 1) with data available since 1979. The data could be extended from 1992 to 1979 by using results provided by Kahru & Elmgren (2014), transforming corresponding indicator variables using a linear model between the two data sets (Anttila et al. 2018).
The shift-detection method was also applied to detect biomass threshold values in the Bothnian Sea, where cyanobacterial blooms did not occur regularly before the late 1990s. In the remaining assessment units where biomass data was not available from a time period before the vast increase of cyanobacterial blooms, threshold values were estimated using the quartile expressing lowest biomass.