Assessment Protocol

The average chlorophyll-a concentration in open sea assessment units is a combined estimate of as many as three types of data (depending on availability, applicability and regional agreement): 1) in-situ measurements 2) Earth Observation (EO) remote sensing satellite data, and 3) FerryBox data. These data are combined as annual averages, applying weighting based on data availability and methodological confidence. The indicator specifics are presented in Assessment protocol table 1.

More information is found in the eutrophication assessment manual.

 

Assessment protocol table 1. Specifications of the core indicator chlorophyll-a.

table1_assessm.jpg 


The in-situ chlorophyll-a data (1) is generated via samples from laboratory extracted and analyzed material, as explained in the HELCOM COMBINE manual. Measurements made at the depth of 0 – 10 m from the surface are used in the assessment.

The satellite-based EO-dataset (2) for 2011 was calculated at SYKE using the ENVISAT/MERIS instrument observations with FUB bio-optical model (Schroeder et al., 2007). The accuracy of the bio-optical algorithm to determine chlorophyll-a concentrations has been validated against ICES monitoring station dataset during HELCOM EUTRO-OPER-project. The EO chl-a values for the surface layer depends on the transparency of the water. Cloudy areas have been removed from the dataset. The data was reported as daily statistics of 20K grid cells (Assessment protocol figure 1).

Information based on flow-through system onboard ferrylines (FerryBox data, 3) was collected and validated by SYKE, and was reported to ICES as daily averages in 20K spatial grids (Fig. 1). As selected ferries operating on the Baltic Sea are the platform for FerryBox flow-through systems and only specific routes are followed (https://www.ferrybox.com/routes_data/routes/baltic_sea/index.php.en) then data availability is not evenly distributed across all HELCOM sub-basins. To remove possible spatial bias, which might be considerable in areas with spatial sampling gradients, we suggest that the Ferrybox-based chla estimate is corrected to represent the entire area. This correction is done at each HELCOM sub-basin, based on a longer-term reference data, which was achieved using remote sensing MERIS estimates from 2002-2011. The correction is done separately for each year within the assessment period, according to the following formula:

formula.png, where                                 

Fcorr is the corrected Ferrybox chla estimate in a HELCOM sub-basin,
n is the number of grid cells in the HELCOM sub-basin,
refi is the grid cell (geometrical) average for the reference data (MERIS 2002-2011, Figure 2),
refave is the sub-basin avegage of the reference data and
Fi is the (geometrical) average Ferrybox chla estimate for grid cell 'i'.

Figure1_assessm.png

Assessment protocol figure 1. Earth observation data are reported as 20K grid cells.

 

In coastal areas the indicator is assessed using comparable indicators developed nationally for the purposes of assessments under the EU Water Framework Directive and data can be derived from different seasons (see Results table 2).

 

Assessment units

The core indicator is applicable in the 17 open sea assessment units (exceeding one nautical mile seawards from the baseline)

In the coastal units the indicator is assessed using comparable indicators developed nationally for the purposes of assessments under the EU Water Framework Directive.

The assessment units are defined in the HELCOM Monitoring and Assessment Strategy Annex 4.

 

Further work required

The use of remote sensing and ship-of-opportunity data for estimating should be developed further, with the aim of extending the temporal coverage of satellite data and spatial coverage of ferry-box data.