Indicator Assessment

Nutrients in transitional, coastal and marine waters

Indicator Assessment
Prod-ID: IND-14-en
  Also known as: CSI 021 , MAR 005
Created 04 Nov 2014 Published 03 Mar 2015 Last modified 09 Apr 2019
21 min read
This is an old version, kept for reference only.

Go to latest version
This page was archived on 09 Apr 2019 with reason: Other (New version data-and-maps/indicators/nutrients-in-transitional-coastal-and-4/assessment was published)

Between 1985 and 2012, most stations in European Seas that reported to the EEA showed no change in trends of concentrations of Dissolved Inorganic Nitrogen (DIN) or orthophosphate. In addition, a decrease in concentrations was observed for 14% and 13% respectively, while only a minority of stations showed an increase.

These trends mostly refer to stations in the northeast Atlantic Ocean and Baltic Sea, however, due to lack of reported data for other regional seas. Available data shows nitrogen and phosphorus concentrations are decreasing in the southern North Sea which is an area with a recognised eutrophication problem. In the Baltic Sea, also affected by eutrophication, nitrogen concentrations are decreasing but phosphate concentrations show an increase at some stations. 

Dissolved inorganic nitrogen concentrations in European seas

Note: The map shows the winter dissolved inorganic nitrogen (DIN) concentrations in European coastal and open waters in 2012. The class boundaries, “high” (red), “moderate” (orange) and “low” (yellow) concentration are determined by the 80/20 percentiles of the DIN data set for the years 2007 to 2012 in each regional sea.

Data source:

Oxidised nitrogen concentrations in European seas

Note: The map shows the winter oxidised nitrogen concentrations in European coastal and open waters in 2012. The class boundaries “high” (red), “moderate” (orange) and “low” (yellow) concentration are determined by the 80/20 percentiles of the data set for the years 2007 to 2012 in each regional sea.

Data source:

Orthophosphate concentrations in European seas

Note: The map shows the winter orthophosphate concentrations in the European coastal and open waters in 2012. The class boundaries “high” (red), “moderate” (orange) and “low” (yellow) concentration are determined by the 80/20 percentiles of the data set for the years 2007-2012 in each regional sea.

Data source:

Trends in dissolved inorganic nitrogen concentrations in European seas

Data sources:
Data sources:

Trends in oxidised nitrogen concentrations in European seas

Data sources:
Data sources:

Trends in orthophosphate concentrations in European seas

Data sources:
Data sources:

Trends per station in orthophosphate concentrations in European seas

Note: Stations showing a statistically significant decrease (green), increase (red) or no trend (grey) of winter orthophosphate concentrations within the period 1985 to 2012. Selected stations must have data, at least, in the period 2007-2012 and must have at least five years data in all.

Data source:

Baltic Sea

In 2012, the highest DIN concentrations (> 22.5 µmol/l) were predominantly observed in coastal waters of Germany, Poland, Lithuania and Russia (Figure 1). High oxidised nitrogen concentrations (> 19 µmol/l) coincided with locations with high DIN (Figure 2). Low concentrations of DIN (< 5.4 µmol/l) were measured along the Swedish coast and in the Baltic Proper. Areas with high orthophosphate concentrations (> 1.1 µmol/l) included the Gulf of Finland and certain coastal locations along the German and Polish coasts (Figure 3). Low phosphate concentrations (< 0.6 µmol/l) were commonly observed in the Baltic Proper and Gulf of Bothnia.

Between 1985 and 2012, DIN concentrations decreased in 7% (11) of monitoring stations and increased in 3% (5) of monitoring stations (Figures 4 & 7). When oxidised nitrogen (OxN) is considered, 7% (12) of monitoring stations showed a decreasing trend, as opposed to an increasing trend shown in 2% (2) of stations (Figures 5 & 8). In the majority of the stations (90%), no significant trend was observed for both DIN and OxN. The direct comparison of DIN and OxN trends is not possible since the number of stations for each variable differs. However, decreasing nitrogen trends were mainly detected in open water stations, Swedish and Danish coastal waters in the western Baltic Sea. Increasing DIN trends were observed in some coastal waters in the Gulf of Finland mainly.

Orthophosphate concentrations decreased in 4% (6) of stations and increased in 9% (15) of stations (Figure 6). Increasing orthophosphate trends were mainly detected in coastal waters. Decreasing trends were observed mainly in the western Baltic Sea along the Danish and German coast (Figure 9).

Greater North Sea

In the Greater North Sea, the highest winter concentrations of DIN (> 64 µmol/l) and orthophosphate (> 1.3 µmol/l) in 2012 were observed in transitional and coastal waters along the Belgian, Dutch and German coast (Figures 1 & 3). Open sea stations showed predominantly low DIN (< 9.2 µmol/l) and OxN (< 8 µmol/l) concentrations (Figures 1 & 2).

Long term (> 10 years) time series indicate that DIN concentrations are declining. In fact, a decreasing trend was observed in 27% (26) of stations of the North Sea, mainly located in transitional and coastal waters along the Belgian, Dutch and German coast. None of the stations showed an increasing trend (Figures 4 & 7). A similar trend was observed for orthophosphate, which decreased in 29% (34) stations of the Greater North Sea (Figures 5 & 8), predominantly in transitional and coastal waters of Belgium, the Netherlands and Germany. Increasing trends in orthophosphate were only recorded in one station. This positive development in nutrient reduction, in particular in phosphorus, can be attributed to improved waste water treatment, which led to a significant reduction of phosphorus loading in most North Sea countries in the period 1985 to 2005 (OSPAR 2008). 

Celtic Seas, Bay of Biscay and the Iberian coast

In 2012, measurements of DIN (and OxN) concentrations were generally limited to Irish and British coastal waters, and some data for the coast of Spain (Figures 1 & 2). High concentrations of DIN (> 59 µmol/l) were observed in coastal waters of Ireland and Great Britain, whereas the concentrations in the coastal stations of Spain were mainly classified as low (< 7 µmol/l).

The coverage of coastal stations was much higher for orthophosphate than for nitrogen, in particular along the Bay of Biscay (Figure 3). Most orthophosphate concentrations were classified as low (< 0.4 µmol/l) or medium (0.4 – 1.1 µmol/l), with the exception of a few Irish coastal stations, in which the orthophosphate concentrations were found to be high (> 1.1 µmol/l).

Not enough time series data was available to evaluate trends in DIN concentrations in Atlantic waters. For oxidised nitrogen, no trends were observed at stations in Ireland. As for orthophosphate, no remarkable trends were observed except for a decreasing trend in three out of 24 French stations in the Bay of Biscay (Figures 6 & 9). 

Mediterranean Sea (combined assessment)

Only a limited number of measurements were reported by Cyprus, Spain, Croatia and Slovenia.  For this reason, the assessments for the regional seas within the Mediterranean sea were combined. Nutrient concentrations in the open Mediterranean Sea are extremely low and eutrophication is only observed in some coastal waters (UNEP 2007). All of the DIN measurements were classified either as low (< 2.3 μmol/l) or medium (2.3-27 μmol/l) (Figure 1). High orthophosphate concentrations (> 0.5 μmol/l) were observed in two coastal stations in Spain, close to the Strait of Gibraltar (Figure 3).

There were only a few notable changes in nutrient concentrations as most stations showed no significant trends (Figure 3,4 & 7,8). However, the evaluation of trends was hampered by the limited number of measurements.

Black Sea

No data for winter means of nutrients in 2012 were available for the Black Sea. Due to lack of data, it was not possible to perform trend analyses.


  • HELCOM, 2009. Eutrophication in the Baltic Sea – An integrated thematic assessment of the effects of nutrient enrichment and eutrophication in the Baltic Sea region. Balt. Sea Environ. Proc. No. 115B.
  • OSPAR, 2008. Eutrophication Status of the OSPAR Maritime Area. Second OSPAR Integrated Report. Publication Number: 372/2008

Supporting information

Indicator definition

This indicator illustrates the levels and trends in winter means of dissolved inorganic nitrogen (nitrate + nitrite + ammonium), oxidised nitrogen (nitrate + nitrite) and phosphate concentrations (micromol/l) in the regional seas of Europe.


Concentrations in micromol/l


Policy context and targets

Context description

Measures to reduce the adverse effects of excess anthropogenic inputs of nutrients and to protect the coastal and marine environment are being taken as a result of various initiatives at all levels - global, European, regional (i.e. through Regional Sea Conventions and/or regional Ministerial Conferences) and national.

There are a number of EU Directives aimed at reducing the loads and impacts of nutrients, including the Nitrates Directive (91/676/EEC) - aimed at the protection of waters against pollution caused by nitrates from agricultural sources; the Urban Waste Water Treatment Directive (91/271/EEC) - aimed at reducing pollution from sewage treatment works and from certain industries; the Integrated Pollution Prevention and Control Directive (96/61/EEC)  - aimed at controlling and preventing pollution of water from industry; the Water Framework Directive (2000/60/EC) - which requires the achievement of good ecological status or good ecological potential of transitional and coastal waters across the EU by 2015; and the Marine Strategy Framework Directive (2008/56/EC) which  requires the achievement or maintenance of good environmental status in European seas by 2020 at the latest, through the adoption of national marine strategies based on 11 qualitative descriptors. The Ecological Descriptor 5 is on Eutrophication.

Measures also arise from international initiatives and policies including: the UN Global Programme of Action for the Protection of the Marine environment against Land-Based Activities; the Mediterranean Action Plan (MAP) 1975; the Helsinki Convention 1992 (HELCOM); the OSPAR Convention 1998; and the Black Sea Environmental Programme (BSEP).


The most relevant EU policy target with regard to chlorophyll concentrations is from the Water Framework Directive (WFD), which aims to reach good ecological status of all EU surface waters by 2015. Member States have defined water-type specific environmental standards to support the achievement of good ecological status. As natural and background concentrations of nutrients vary between and within the regional seas, and between types of coastal water bodies, nutrient targets or thresholds for achieving good ecological status have to be determined while taking into account local conditions.

Within the scope of the Marine Strategy Framework Directive, nutrient levels (nutrient concentrations in the water column and nutrient ratios for nitrogen, phosphorus and silica, where appropriate) are the relevant criteria and indicators in marine waters under the Goof Environmental Status (GES) Descriptor 5: Human-induced eutrophication. The aim of the MSFD is to reach or maintain GES of the marine environment by 2020. The assessment of eutrophication in marine waters needs to combine information on nutrient levels as well as a range of ecologically relevant primary and secondary effects, taking into account relevant temporal scales. The nutrient targets and thresholds for achieving good environmental status have not been defined yet.

Related policy documents



Methodology for indicator calculation

The data used in this indicator is part of the WISE - State of the Environment (SoE) data, available in Waterbase - TCM (Transitional, Coastal and Marine) waters. Waterbase is the generic name given to the EEA´s database on the status, quality and quantity of Europe´s water resources. Waterbase – TCM waters contains data collected from both EEA member countries (i.e. belonging to the EIONET) and from the Regional Seas Conventions through the WISE-SoE TCM data collection process. The resulting WISE SoE TCM dataset is therefore made of sub-samples of national data assembled for the purpose of providing comparable indicators of state, pressures and impacts of transitional, coastal and marine waters (TCM-data) on a Europe-wide scale.

Consistent time series are used as the basis for assessment of the development over time. The trend analyses are based on time series from 1985 onwards. Stations with  data in, at least, the last six years (2007 or later), and five or more years in the period since 1985 are selected. For nitrogen, dissolved inorganic nitrogen (DIN – nitrate + nitrite + ammonium) and oxidised nitrogen (OxN - nitrate + nitrite) are both used. In the case of gaps in nitrite data, only ammonium and nitrate were used to complete the time series.

In previous assessments, the description of nitrogen was limited to oxidised nitrogen (nitrite + nitrate) and did not include ammonium (NH4+), which is another important inorganic nitrogen compound. This approach deviated from the common practice in most Regional Sea Conventions and that described in the WFD and MSFD, where all dissolved inorganic nitrogen compounds (described as DIN=nitrite + nitrate + ammonium) should be taken into consideration. In many cases, ammonium is a substantial fraction of winter average DIN and a description of oxidised nitrogen (nitrite + nitrate) alone may therefore lead to a biased assessment result. For the sake of comparison with previous assessments, both DIN and OxN are considered in this assessment.

 The winter period is defined as follows:

  • January, February and March for stations east of longitude 15 degrees (Bornholm) in the Baltic Sea
  • January and February for all other stations


Primary aggregation

The primary aggregation consists of:

  • Identifying stations and assigning them to countries and sea regions.
  • Creating statistical estimates for each combination of station and year.

Geographical classification: Sea region, coastal/offshore

All geographical positions defined in the data (i.e. in stations) are assigned to Europe´s regional seas by coordinates and used in the aggregation process for different determinants. The stations are then further classified as coastal or open water (>20 km from coast) by checking them against the coastal contour. Open waters stations – off-shore - are distinguished per regional sea, whereas coastal stations are further attributed to country. These classifications are done in ArcGIS.

Eionet stations

TCM data reported directly from countries are assigned to station identifiers that are listed with coordinates.

Marine convention data from ICES

For the data reported through ICES, there are no consistent station identifiers available in the reported data but only geographical positions (latitude/longitude). The reported coordinates for what is intended to be the same station may vary between sampling visits because the exact sampling position is recorded, not the target position. Identifying a station from its real sampling position may fragment time series too much. Therefore, for open waters (>20 km from land), coordinates are rounded to two decimal points. This is used to create stations (i.e. time series) with station names derived from rounded coordinates. The station coordinates are the average in the sampling visits to the station rather then the rounded coordinates. This ensures that, in cases where most observations are in a tight cluster within the rounding area, a position within the cluster is used. For the coastal ICES stations, there may be some overlap with Eionet stations. In coastal stations rounding coordinates to two decimal points may be too much (about 500m to 1km). However, the rounding is also done for coastal stations but the grouping of observations to rounded coordinates is done separately only within observations from each country and the originator country is listed. Note that these stations are not necessarily close to the coast of the originator country.

Many countries have made measurements over large areas, including some observations fairly close to the coast of other countries, although probably not normally within territorial waters of other countries. This means that, at least for open waters, assigning data to originating country may not necessarily reflect geographical location. Duplicates between data reported through ICES or from the Eionet directly may occur. A visual inspection of coastal data (< 20 km from shoreline) is performed to identify those issues and correct them where possible (namely through feedback with the originator country(ies).

Statistical aggregation per station and year

The aggregation is done in two- or three-stage query sequences, which include:

  • Selecting season (month) and depth;
  • If needed, building a cross-table with determinants in columns, and water samples in rows, and deriving composite determinants from that;
  • Aggregating over depth for each combination of station and date; and
  • Aggregating over dates within each combination of station and year. 

The basic data consists of two tables:

Measurement values table
WaterbaseID (Country and Station)
Date (Year, Month and Day)
Determinant, with Determinant codes "Ammonium", "Nitrate", "Nitrite", "DIN", "Total oxidised nitrogen" and "Orthophosphate".


Stations table
Unique identifier: data provider, Country and Station ID
Sea region (Atlantic, North Sea, Baltic, Mediterranean and the Black Sea)

The two tables are combined in a query which joins data to stations, linked by WaterbaseID, and including Country Code and Sea Region (used in Selection Criteria below). This query (or a table made from it) is used in the Aggregate queries.

Description of specific aggregation query sequencesNitrate and Phosphate
Step 1

Crosstable query, with determinands "Ammonium", "Nitrate", "Nitrite", "DIN", "Total oxidised nitrogen" and "Orthophosphate" as columns, and row heading Sea Region, WaterbaseID, Year, Month, and SampleDepth.
Include data for:
SampleDepth <=10 m and 
Month = 1,2,3 (Jan. - Mar.) for stations east of longitude 15 degrees (Bornholm) in the Baltic Sea
Month = 1,2 (Jan.- Feb.) for all other stations.

Step 2

For each combination of WaterbaseID*Year*Month*Day, calculate [Total Oxidised Nitrogen] and [DIN]: Calculate best possible estimate of nitrate including nitrite 

Aggregate arithmetic mean of Oxidised Nitrogen and Orthophosphate over depths.

Step 3

For each combination of WaterbaseID *Year,

calculate the arithmetic mean over the depth averages from Step 2.
Export result to Aggregate database as table 't_Base_Metadata_N_and_P'



Concentrations from the most recent year available (2012) are presented on a map, where concentrations are classified as low, moderate or high. Low concentrations are defined as concentrations smaller than the 20-percentile value of concentrations within the specific regional sea in the last six years (i.e. 2007-2012). High concentrations are concentrations higher than the 80-percentile value of concentrations within the regional sea in the last six years (i.e. 2007-2012). All other concentrations are classified as moderate. This classification helps to identify areas of low and high concentrations and is based on six-year percentiles, unlike previous assessments, which only considered the percentile values within a regional sea based on data from that specific year.


Trend analysis

Trend analysis was carried out for each station in a region having data in, at least, the last six years (2007 or later), and at least five or more years in the period since 1985. Trend detection for each time series was done with the Mann-Kendall Statistics using a two-sided test with a significance level of 5% (Sokal & Rohlf 1995).

In the presentation of the results, a distinction is made between trends based on relatively short time series (≤ 10 years) and longer time series (> 10 years).

A nutrient-salinity gradient is commonly observed along the freshwater-seawater continuum in transitional, coastal and marine water bodies. The variation of nutrient concentration with salinity is commonly represented in “mixing curves”, with elevated nutrient concentrations at the freshwater end decreasing towards marine waters. At some stations in this salinity gradient, year-to-year variations in salinity may result in large variations in nutrient concentrations. For stations where interannual variations in nutrient concentrations were correlated with changes in salinity, data were corrected for the salinity effect before carrying out a trend analysis.

The Mann-Kendall method is a non-parametric test suggested by Mann (1945) and has been extensively used for environmental time series (Helsel and Hirsch, 2002; Hipel and McLeod, 2005). Mann-Kendall is a test for monotonic trends in a time series y(x), which in this analysis is nutrient concentration (y) as a function of year (x). The test is based on Kendall's rank correlation, which measures the strength of monotonic association between the vectors x and y. In the case of no ties in the x and y variables, Kendall's rank correlation coefficient, tau, may be expressed as tau=S/D where S = sum {i<j} (sign(x[j]-x[i])*sign(y[j]-y[i])) and D = n(n-1)/2. S is called the score and D, the denominator, is the maximum possible value of S. The p-value of tau is computed by an algorithm given by Best and Gipps (1974). The tests reported here are two-sided (testing for both increasing and decreasing trends). Data series with p-value < 0.05 are reported as significantly increasing or decreasing. The test analyses only the direction and significance of the change, not the size of the change.

The Mann-Kendall test is a robust and accepted approach. Due to the multiple trend analyses, approximately 5% of the conducted tests will turn out significant (identify a trend) if in fact there is no trend. The accuracy at the regional level is largely influenced by the number of stations for which data is available.

Methodology for gap filling

For nitrogen, dissolved inorganic nitrogen (DIN – nitrate + nitrite + ammonium) and oxidised nitrogen (OxN - nitrate + nitrite) are both used. In the case of gaps in nitrite data, only ammonium and nitrate were used to complete the time series assuming the nitrite fraction was negligible.

Methodology references



Methodology uncertainty

The Mann-Kendall test for the detection of trends is a robust and accepted approach. However, due to the multiple trend analyses, approximately 5% of the tests conducted will turn out significant if, in fact, there is no trend. Also, the accuracy at the regional level is largely influenced by the number of stations for which data is available.

There are also a number of uncertainties related to temporal and spatial use of the data. Currently, the winter period is defined as January and February for all stations except for stations east of longitude 15 degrees (Bornholm) in the Baltic Sea. However, this definition may be too broad to reflect the climatic differences across the European sea regions. For example, for the Black Sea, it is suggested to also consider spring concentrations due to the nutrient enrichment of coastal waters as a result of increased riverine inputs (BSC, 2010). 

Moreover, two types of geographical aggregation are performed in the current methodology, based on the Country Code and Sea Region. In both cases, differences in physical, chemical and biological characteristics between sampling stations are not taken into account. Measured nutrient concentrations should be related to natural background values that reflect spatial/geographical differences. Furthermore, data collected over the different years is obtained from different laboratories, possibly following different methodologies and it is combined in the same trend analysis. This might also influence the results.

Data sets uncertainty

Data for this assessment is still scarce, considering the large spatial and temporal variations inherent in European transitional, coastal and marine waters. Long stretches of European coastal waters are not covered in the analysis due to lack of data or sufficiently long and recent time series. 

Rationale uncertainty

Due to variations in freshwater discharges, the hydro-geographic variability of the coastal zone and internal cycling processes, trends in nutrient concentrations, as such, cannot be directly related to measures taken in nearby river basins. However, overall trends reflect the effects of measures to reduce nutrient pollution.



Data sources

Other info

DPSIR: State
Typology: Descriptive indicator (Type A - What is happening to the environment and to humans?)
Indicator codes
  • CSI 021
  • MAR 005
Frequency of updates
Updates are scheduled every 2 years
EEA Contact Info


Geographic coverage

Temporal coverage