EU Ecosystem Assessment - Freshwater Ecosystem

Collection: MAES : Mapping and Assessment of Ecosystems and their Services 


List of indicators included in the dataset: 1) River, Lakes and Freshwater ecosystem extent; 2) Domestic waste emissions to the environment; 3) Nitrogen, Phosphorous, and Biochemical Oxygen Demand concentration in rivers below alert thresholds; 4) Frequency of low flow alteration - Q10; 5) Water Exploitation Index (Consumption) – WEIc; 6) Land cover in riparian areas; 7) Morphological Alteration of freshwater habitat.

1) River, Lakes and freshwater ecosystem extent Freshwater ecosystems comprise aquatic habitats of rivers and lakes, but their functionality depends on the exchanges with the land. Riparian areas and floodplains can also be seen as complementary to rivers and lakes as part of freshwater ecosystems, and provides many of their linked ecosystem services. Rivers are elongated landscape forms whose limited width precludes effective detection through satellite images (Pekel et al. 2016). The extent of freshwater ecosystems was derived combining several datasets:

1- Rivers and lakes were defined by the network of the Catchment Characterisation Model (CCM; De Jager & Vogt 2007). 2- In order to restrict the analysis to freshwaters, brackish or salty water bodies as classified by Corine CLC 2012 into classes 37-39 and 42-44), were removed. The extent of rivers is expressed in km of length per administrative unit NUTS2; the extent of lakes is quantified by surface area (km2). 3- The area of aquatic ecosystems (river and lakes, km2) was quantified by buffering CCM river segments with a 5m buffer on each side, i.e. assuming a constant river cross section of 10m everywhere. This is a simplification borne from the lack of data. Rivers and lakes were merged into one freshwater aquatic ecosystem extent layer (FreshwaterExt_RivLakes) 4- Riparian land is the zone where most exchanges between aquatic habitats and surrounding land take place. Riparian land was defined according to Weissteiner et al. (2016). Riparian land was merged with that of Rivers and Lakes (FreshwaterExt_RivLakesRip). 5- Potentially flooded areas receive waters from rivers in the occasion of flood events. The exchanges between rivers and lakes with flooded areas are more episodic in nature than for riparian land, yet this zone is very important to assess some freshwater ecosystem services, like protection from flood hazards. The potential flood prone areas were defined as the area that can be flooded by a 100-year return event (EEA, 2020). The potential flood prone areas were merged with River, Lakes and Riparian areas (FreshwaterExt_RivLakesRipFlood). 6- All datasets were dissolved as polygons for NUTS2 administrative level. Temporal coverage: 2012.

2) Domestic waste emissions to the environment Pollutant emissions from domestic sources impact freshwaters nutrient balance. Domestic waste comprises emissions from point sources, i.e. direct discharges to river network form wastewater treatment plans and sewerage systems, and from diffuse sources, i.e. disconnected dwellings and independent appropriate systems. Population Equivalent (PE) reported in the Urban Waste Water Directive database (v5 with reference year 2014, revised for data gap fillings) were used to assess annual emissions of biochemical oxygen demand (BOD), nitrogen and phosphorous loads (t/y). Removal efficiencies depending on treatment levels were applied before releases to the environment. It should be noted that emissions from disconnected sources would be further attenuated before reaching the river network; however, this attenuation has not been considered for these reported domestic emission indicators. The dataset reports mean annual domestic emissions of total nitrogen (N), total phosphorous (P), and 5-days biochemical oxygen demand (BOD) loads per km2 (t/y/km2) for EU28 at NUTS2 scale. Temporal coverage: 2014-2015

3) Nitrogen, Phosphorus, and Biochemical Oxygen Demand concentration in rivers below alert thresholds High nutrient concentrations in rivers are conducive to eutrophication. Mean annual nutrient concentrations (mg/L) of nitrogen, phosphorous and Biochemical Oxygen Demand (BOD5) were obtained with conceptual pollution source-fate models. Specifically, nitrogen and phosphorous concentrations were modelled with GREEN (Grizzetti et al., 2012) as mean annual values for the 2005-2012 hydrological conditions. BOD5 was modelled as in Vigiak et al. (2019) for 2008-2012 mean hydrological conditions. The indicators are presented as fraction of river length (in km/km, aggregated at NUTS2 level) where concentrations are below the following thresholds: total nitrogen <= 4 mg/L; total phosphorous <= 0.1 mg/L, and BOD5 <= 5 mg/L as per Pistocchi et al. (2019). Temporal coverage: 2010

4) Frequency of low flow alteration - Q10 Heavy water abstractions may reduce streamflow in aquatic freshwater systems to levels that threaten biodiversity survival and ecosystem services delivery, to the point that some countries forbid abstractions when streamflow is below given thresholds. We used as reference the 10th percentile of streamflow that would occur in natural conditions (i.e the level below which streamflow would occur on average 36.5 days a year in the absence of any abstraction and flow regulations), and calculated the frequency with which streamflow is below the reference in current hydrological conditions. The indicator is calculated with Lisflood 2.0 model (Bisselink et al., 2018) for historical records of 2000-2018 for current conditions. Natural conditions are simulated by taking out water abstractions and flow regulations (reservoirs). The indicator Q10 is calculated at pixel size of 5 x 5 km2. A frequency below 0.1 means that the natural low flow occurrence is respected, whereas values above the 0.10 thresholds indicate increasing alteration of low flows. Currently in the EU28 territory 61% of freshwaters are affected by some degree of alteration (i.e. the frequency of streamflow above the Q10 threshold is less than 90% of time). In 5% of EU28 territory low flow alteration is extreme, below the natural 10th percentile for more than 50% of time. Temporal coverage: Average 2000-2018

5) Water Exploitation Index (Consumption) – WEIc Water abstractions pose heavy pressure on aquatic freshwater ecosystems, especially where water availability is limited. The Water Exploitation Index on consumption (WEIc) is the ratio of net water consumption divided by the freshwater resources of a region. Water consumption considers net abstractions, i.e. total abstractions less the amount of water that returns to the system after usage. WEIc above 0.20 indicates water stress, and WEIc above 0.4 indicates severe water stress. The indicator is calculated with the Lisflood 2.0 model (Bisselink et al., 2018). Water abstractions and consumption rates are defined for domestic, industrial, energy and cooling, livestock and irrigation. Freshwater resources taken into account include locally generated runoff, inflowing surface waters, lakes and reservoirs storage and outflow, and groundwater recharge (Bisselink et al. 2018). For the purposes of MAES, WEIc is calculated for sub-river basin districts within an EU Member State as the average for 2000-2018. Currently 8% of EU28 suffers of water stress (WEIc => 0.20). Temporal coverage: Average 2000-2018

6) Land cover in riparian areas Riparian areas exert important regulatory services like water purification and provide habitats to freshwater biota. Encroachment of these areas by anthropogenic activities limit these functionalities (Pistocchi et al., 2015; 2017; 2018). Riparian land was identified according to Weissteiner et al. (2016). Land cover was taken from CORINE CLC status (CLC2000, CLC2006, CLC2012 V18; CLC2018 release V20b2; accessed 15/06/2019). The share of artificial land use is taken from CORINE Land Cover level 1 category 1. The share of agricultural land use is taken from CORINE Land Cover level 1 category 2. The share of natural land is taken from CORINE Land Cover level 1 categories 3 and 4. Land use shares of riparian land are presented at NUTS2 level.

Temporal coverage: 2000; 2006; 2012; 2018

7) Morphological Alteration of freshwater habitat Indicators of morphological alteration of freshwater habitats were assessed by the presence of structures that may interrupt the lateral and longitudinal connectivity of rivers. The presence of infrastructures in riparian areas can be considered a proxy for human pressures on the aquatic habitats, disturbing the natural setting and potentially altering lateral connectivity. The presence of dams, locks and other barriers along streams reduces the fluxes of water, materials and species along the stream network, interrupting the longitudinal connectivity of aquatic habitats. Barriers limit sediment supply and alter natural flow regime, and therefore may induce river bed aggradation and alter channel-alluvium exchanges (Pistocchi et al., 2018). Three indicators of morphological alteration are proposed: 1- InfrDensity: the density of infrastructures in riparian land. Linear infrastructures (roads and railways) in riparian areas were mapped on the basis of the OpenStreetMap road and railway segments accessed in 2016. The extent of riparian land was mapped by Weissteiner et al. (2016). Density is expressed in km of linear infrastructure per km2 of riparian land. 2- DamPerc: the fraction of mean annual streamflow intercepted by dams/barriers, giving an indication of longitudinal hydrological disconnectivity; and 3- FreeL_fraction: the fraction of the stream network that is dams-free, i.e. the ratio of the stream length theoretically accessible between barriers divided by the total length of stream network, providing an indication of longitudinal morphological disconnectivity. For each river segment (mapped with CCM tessellation), the total length of stream network to which it belongs (from headwater to sea outlet) is calculated. Then, the sum of lengths of segments that are accessible in the presence of dams is calculated. A segment is accessible if there is no dam in between. The indicator is the ratio of the total accessible length over the total length of the segment stream network and represents the share of naturally available habitat. The analysis was conducted at 1 km resolution. Dams, weirs and locks were delineated enlarging the EuroRegionalMap v6.04, including some 3000 dams with a reservoir of at least 0.4 km2 surface, with other sources. The dataset can be considered valid for 2015, however mapped barriers are incomplete. Further, presence of mitigation measures could not be considered. For the purposes of MAES, the indicators were aggregated at NUTS2 level. Temporal coverage: 2015




  • Bruna Grizzetti
  • Arie de Roo
  • Olga Vigiak
  • Alberto Pistocchi
  • Angel Udias Moinelo
  • Alberto Aloe
  • Michela Zanni
  • Chiara Dorati
  • Christof Weissteiner
  • Faycal Bouraoui
  • Bernard Bisselink
  • Marko Adamovic
  • Susann Guenther
  • Lorenzo Mentaschi
  • Trine Christiansen
  • Lidjia Globevnik
  • Jeroen Bernhard
  • Emiliano Gelati

How to cite

Grizzetti, Bruna; de Roo , Arie; Vigiak, Olga; Pistocchi, Alberto; Udias Moinelo, Angel ; Aloe, Alberto; Zanni, Michela; Dorati, Chiara; Weissteiner, Christof; Bouraoui, Faycal; Bisselink, Bernard; Adamovic, Marko; Guenther, Susann; Mentaschi, Lorenzo; Christiansen, Trine; Globevnik, Lidjia; Bernhard, Jeroen; Gelati, Emiliano (2021): EU Ecosystem Assessment - Freshwater Ecosystem. European Commission, Joint Research Centre (JRC) [Dataset] PID:


freshwater, ecosystem, rivers, lakes, riparian zone, potentially flooded areas, Nitrogen, Phosphorus, Biochemical Oxygen Demand, TN, TP, BOD, domestic waste, streamflow, LISFLOOD, ecosystem pressure, hydromorphological alteration, environmental quality, rivers and lakes

Data access

Download Freshwater Ecosystem Indicators - EU Ecosystem Assessment

Geographic areas

European Union

Additional information

Published by
European Commission, Joint Research Centre
Created date
Modified date
Issued date
Landing page 
Data theme(s)
Environment, Health, Population and society
Update frequency
01 Oct 2021: 1 visits