Indicator Assessment

Imperviousness and imperviousness change

Indicator Assessment
Prod-ID: IND-368-en
  Also known as: LSI 002
Publicerad 2017-12-04 Senast ändrad 2021-05-11
21 min read
This page was archived on 2020-03-25 with reason: Other (New version data-and-maps/indicators/imperviousness-change-2/assessment was published)
  • Between 2009 and 2012, soil sealing (or imperviousness) increased in most EEA-39 countries by a total of 2 051 km2. This corresponds to 0.0356 % of the total EEA-39 area or an annual average increase of 683 km2. Overall, this rate of increase is lower than that documented in the 2006-2009 products, although this can be put down to a general overestimation of sealing increase in the 2006 and 2009 datasets. 

  • In 2012, the percentage of a countries' total area that was sealed also varied greatly, with values ranging from 0.14 % to 16.27 %. The highest sealing values, as a percentage of country area, occurred in small countries with high population densities, while the lowest sealing values could be found in large countries with low population densities. The most problematic situation occurs in countries where there is already a high percentage of sealing and where the annual rate of increase relative to country area is high.


Percentage soil sealing by country

Bar chart

Imperviousness density in 2012

Note: The map shows the density of soil sealing in 2012, based on a 10 km2 reference grid. Green and light orange colors show areas with no or very limited sealing, while red and dark red colors show highly to fully sealed grid cells (mainly urban areas).

Data source:

Annual change in soil sealing between 2009-2012, relative to sealed area in 2009


Annual average change in soil sealing, 2009-2012, relative to country area

Bar chart

Average annual change in soil sealing, 2009-2012

Note: The map shows the percentage of the average annual change in soil sealing for each of the rectangular 10 km x 10 km grid cells, over the 2009-2012 period

Data source:

Absolute increase of sealing between 2009 and 2012

Between 2009 and 2012, the extent of soil sealing in the EEA-39 countries increased by 2 051 km2 to a total area of 117.722 km2 (or 2.044 % of the EEA-39 area), according to imperviousness data. This corresponds to 0.0346 % of the total EEA-39 area or an annual average increase of 683 km2. Overall, the rate of increase is lower than that documented in the 2006-2009 products, although this can likely be put down to a general overestimation of sealing increase and absolute sealing levels between 2006 and 2009[1].  A fully harmonised comparison of the sealing dynamics for this period will only become available in 2018, with a full reprocessing of the time series. In general, it is important to keep in mind that while the time-series and the captured rates of sealing and sealing change are being harmonised and improved, real sealing levels in most countries are above those captured in the imperviousness products. The imperviousness change and status products are reliably capturing the spatial pattern and magnitude of sealing, and the general trends in sealing change, but due to the pixel-size of the satellite imagery used (20 m), very small-scale sealed surfaces are often not captured. In addition, the product deliberately does not 'burn in' any additional datasets (e.g. road network data), because it would reflect directly any errors and inconsistencies in update frequency and quality of the additional input data. This needs to be considered if the sealed area from the imperviousness data is compared to sealing statistics available in some countries, that are based on very precise and constantly updated cadastral data.

Percentage of country area sealed

While on average, 2.76 % of the total area covered by the countries was sealed in 2009[2], this increased to 2.87 % in 2012. See Figure 1 for a comparison of country averages between 2009 and 2012, and Figure 2 for a map of sealing density in 2012, based on a 10 km grid.

In 2012, three groups of countries could be identified: 

a) Countries with a very low sealed area as a percentage of total area (i.e. <1.0 %), e.g. Iceland (0.14 %), Norway (0.27 %), Sweden (0.51 %) and five other countries. 
b) Countries with medium-high sealing values between 1.0 % and 3.0 %. There are 19 such countries.
c) A considerable number of countries with high sealing values between  3.0% and 8.0 %. There are 12 such countries, including Germany (5.17 %), Liechtenstein (6.29 %), Belgium (7.56 %), the Netherlands (8.07 %) and Malta (16.27 %). See figure 1.

A more detailed spatial pattern of existing sealing in 2012, based on 10 km grid cells, can be seen in Figure 1. The main urban areas that have grid cells with high sealing levels are shown in dark red, while green areas represent grid cells without any sealing detected in the 2012 imperviousness product.

Annual percentage increase in sealing relative to sealed area in 2009

Sealing increased in most countries between 2009 and 2012. The average annual percentage increase for the period (relative to the sealed area in 2009) was  0.52 % across the EEA-39, with annual values ranging from only very slight increases close to 0 % for Lithuania to 15 % for Liechtenstein. See Figure 3 for yearly increases by country. However, it should be noted that in countries with little relative and absolute sealing (e.g. Norway), even relatively modest (absolute) increases create high change rates.

Percentage annual increase in sealing area (2009-2012) relative to country area (imperviousness indicator)

The imperviousness indicator shows that, on average, the annual rate of increase amounted to 0.0122 % of the total EEA-39 area for the 2009-2012 period. The annual percentage increase in sealing relative to country area ranges from slightly negative values for Sweden and Malta[3], to 0.0002 % in Iceland and values above 0.03 % for Spain, the Netherlands, Cyprus and Italy. See Figure 4 for a graph with results by country, and Figure 5 for a map with indicator results based on a 10 km grid.

Given that the imperviousness indicator does not capture land cover flows, it is not possible to monitor directly which land cover is mainly affected by an increase in sealing. The indicator can, however, be used together with Corine Land Cover (CLC) to establish in more detail which land cover strata are mainly affected by sealing increases.

On average, most of the sealing increases between 2009 and 2012 occurred in the CLC classes listed below. 

CLC classes with the highest sealing levels in 2012:

111 Continuous urban fabric (71.77 %)

123 Port areas (61.72 %)

121 Industrial or commercial units (52.99 %)

112 Discontinuous urban fabric (36.96 %)

122 Road and rail network and associated land (36.05 %)

124 Airports (23.47 %)

133 Construction sites (13.65 %)

141 Green urban areas (12.36 %)

142 Sport and leisure facilities (10.18 %)

CLC classes with the highest percentage increase in sealing between 2009-2012:

133 Construction sites (1.791 %)

123 Port areas (1.270 %)

122 Road and rail network and associated land (1.158 %)

121 Industrial or commercial units (0.849 %)

132 Dump sites (0.610 %)

131 Mineral extraction sites  (0.373 %)

112 Discontinuous urban fabric (0.335 %)

124 Airports (0.311 %)

111 Continuous urban fabric (0.307 %)




[1] Out of the existing time series of imperviousness data (2006, 2009, 2012), the first two datasets were produced as part of the Geoland2 project. During production of the Copernicus products 2012 and 2009-2012, it was detected that there is likely some level of over-estimation in the absolute 2006 and 2009 sealing levels, and in particular in the sealing increases in already sealed areas between 2006 and 2009. This is the reason that the Copernicus Land Monitoring Service, in the context of producing a 2015 reference year update, is currently re-processing the whole time-series for imperviousness products. More reliable and comparable sealing figures will be available from around early 2018, covering 2006-2009-2012 and 2015.

[2] It should be considered that the figures for sealed area in 2009 are not corrected for commission errors, while the change figures show the real change. This means that in some cases, the 2009 imperviousness status map and derived sealing area or percentages are higher than in 2012, while the change product shows a sealing increase from 2009 to 2012. In other words, areas wrongly classified as sealed in 2009 were not recorded as change for the change product, but not (for now) corrected in the 2009 status product.

[3] The negative values for Malta are likely artifacts in the data and the value for Malta is excluded from the analysis. In the case of Malta, the reason is likely to be a geometric shift problem in the input data, while in case of Sweden, the decreases mapped are very small (66 ha) and within the error margin for such a large country. These decreases are likely caused by some larger areas wrongly identified as decreasing in sealing in the 2009-2012 change product. Overall, we will only have a more reliable and comparable picture of the sealing dynamics from 2006 once a fully reprocessed time series is published in 2018.



Supporting information

Indicator definition

The imperviousness indicator is defined as the yearly average imperviousness change between two reference years, as measured by imperviousness change products. The change is aggregated for a certain reference unit and expressed as relative to the size of that unit (as a percentage). The imperviousness change value for a 100 m raster cell is based on 100 m imperviousness change products. The default reference unit is the country, but the indicator can be aggregated based on different spatial units. For example, for a certain country, an imperviousness indicator value of 0.2 %, means that on average, an additional 0.2 % of this country's area has been sealed annually during the period between the two reference years in question. If above a certain rate of increase (threshold values), this value can be used as a warning sign. The aggregation of imperviousness values to reference units is performed using the LEAC CUBE method.


The unit used for this indicator is the yearly average percentage change in imperviousness relative to the size of the reference unit. This was initially based on the total difference over the three reference years (2006-2009, and now 2009-2012) divided by three, but it will be possible to calculate the indicator for different periods at a later stage (e.g. 2006-2015). It is important to note that the yearly average value is based specifically on the period reported, e.g. 2006-2009 or 2009-2012.


Policy context and targets

Context description

The main policy objective of this indicator is to measure the extent and dynamics (change) of soil sealing, resulting from the development of urban and other artificial land uses.

At the United Nations Conference on Sustainable Development, held in Rio in 2012 (Rio+20), world leaders identified land and soil degradation as a global problem and committed to 'strive to achieve a land degradation neutral world in the context of sustainable development'. At the EU level, the Seventh Environment Action Programme (7th EAP) includes a strong focus on the unsustainable use of land and soil, including explicitly the issue of soil sealing. In this context, the 7th EAP refers to a Commission Staff Working Document with the title 'Guidelines on best practice to limit, mitigate or compensate soil sealing' (SWD/2012/0101).

In addition, land take is explicitly mentioned in chapter 23 of the 7th EAP, stating that:

Every year more than 1 000 km² of land are taken for housing, industry, transport or recreational purposes. Such long-term changes are difficult or costly to reverse, and nearly always involve trade-offs between various social, economic and environmental needs. Environmental considerations including water protection and biodiversity conservation should be integrated into planning decisions relating to land use so that they are made more sustainable, with a view to making progress towards the objective of 'no net land take', by 2050.'

In recognition of the importance of land in safeguarding natural resources, the Commission is considering a Communication on 'land as a resource'.

Other important references can be found in A Sustainable Europe for a Better World: A European Union Strategy for Sustainable Development (COM(2001)264) and the thematic documents related to it, such as the Commission Communication Towards a Thematic Strategy on the Urban Environment (COM(2004)60), Cohesion Policy and Cities: the urban contribution to growth and jobs in the regions (COM(2006)385), Europe 2020 (COM(2010)2020), general provisions on the European Regional Development Fund, the European Social Fund and the Cohesion Fund Council Regulation (EC) No 1083/2006, as well as the concept of territorial cohesion.


Although there are no quantitative targets for soil sealing/imperviousness at European level, different documents reflect the need for better planning to control urban growth and the extension of infrastructures. Policies relating explicitly to land use issues, and especially physical and spatial planning, have, until now, generally been the responsibility of the authorities in Member States. The European Commission's Roadmap to a Resource Efficient Europe (COM(2011) 571) introduces, for the first time, a 'no net land take by 2050' initiative that would imply that all new urbanisation will either occur on brownfields or that any new land take will need to be compensated by reclaiming artificial land.

European policy, although it has no spatial planning responsibility, sets the framing conditions for planning. At the European level, the 1999 European Spatial Development Perspective (ESDP), a non-binding framework that aims to coordinate various European regional policy impacts, already advocates the development of a sustainable, polycentric and balanced urban system with compact cities and a strengthening of the partnerships between urban and rural areas, as well as parity of access to infrastructure and knowledge, and wise management of natural areas and cultural heritage. The 2008 Green Paper on territorial cohesion, and the 2007 EU Territorial Agenda and Action Plan by the Territorial Agenda of the EU and the Action programme for its implementation (COPTA, 2007) build further on the ESDP. Specific, relevant actions in the field of 'Land', in particular are Action 2.1d 'Urban sprawl' and Action 2.2 'Territorial impact of EU policies'.

Demand for new urban areas may be partly satisfied by brownfield remediation. Its environmental advantages are clear: relieving pressure on rural areas and greenfield sites, reducing pollution costs, more efficient energy use and natural resource consumption, facilitating economic diversification and emerging habitat (housing) requirements. Europe has several examples of regional strategies for economic regeneration and brownfield development (The OECD Territorial Outlook 2001). On average land recycling increased steadily between 1990 and 2012 on annual basis, with considerable variation between countries, and within countries. Stronger links between EU urban and soil policies could encourage this further (e.g. following up respective 6th EAP Thematic strategies).

Related policy documents



Methodology for indicator calculation

  • The 100 m imperviousness change product, produced as a deliverable in the 2009 and 2012 updates, is the basis for this indicator. 
  • The indicator for the 2006-2009 period (published in 2015) is fully replaced by this first update for the 2009-2012 period, for reasons detailed (among others) in the section on 'Data sets uncertainty'. Once a fully harmonised and reprocessed time series of imperviousness data is available, the indicator will be updated to cover the whole period from 2006 to 2015.
  • Although imperviousness changes included in the 100 m change product could be summarised directly as certain reference units, the calculation of the indicator is performed by ingesting 100 m imperviousness change values into the EEA LEAC cube in order to create harmonised results with other spatial indicators using a similar methodology. The LEAC cube enables the extraction of statistics based on a system of grid-cells with a 1 km side length (an area of 100 ha). 
  • Reference units, such as countries, NUTS3 regions etc. are already included in the LEAC cube as so called 'dimensions' and are used for the statistical aggregation of imperviousness changes corresponding to required reference units.
  • Yearly average imperviousness change values are calculated by dividing the average imperviousness change values by the number of years between two status years.

Methodology for gap filling

  • Known gaps were caused by image gaps (missing input Earth observation data) in both 2006 and 2009 data. For 2009 gaps, the available 2006 values were used for the 2009 status layer and gaps in 2012 image data were filled with imperviousness data from 2009. This means that in some areas (under image gaps), the real sealing change is not known for 2009 and 2012, and some under-estimation of the real sealing increase (for those areas) is likely (because data from previous reference years are used). This practice of gap-filling with previous reference year data will be discontinued for the 2015 (and future) updates, but it will not be possible to eliminate the effects from the 2009 and 2012 datasets.

Methodology references

No methodology references available.



Methodology uncertainty

  • The methodology for deriving the indicator is simple and introduces very little uncertainty. However, it needs to be fully understood that the yearly averages are only valid for the 3-year reference period under consideration. Possible variation within the 3-year period (for individual years) is currently not captured.

Data sets uncertainty

  • This indicator is based directly on the mapping of soil sealing/imperviousness using Earth observation data of about 20 m spatial resolution. Real sealing will differ from the values derived in this way for various reasons:

    • The spatial resolution of the input imagery (20 m) means that very small sealed surfaces will not be captured, e.g. small buildings and small paved roads, and other sealed surfaces with a very small footprint. This can lead to an underestimation of real sealing.
    • As with all Earth observation derived products, the data contain omission errors (sealed surfaces not detected) and commission errors (areas wrongly classified as sealed). The distribution of these errors depends on the quality of the input data, calibration during production and local differences in spectral contrast (which makes sealing in some locations 'easier' to detect than in other locations, depending on the context).
    • There are some known issues in the 2006 and 2009 imperviousness data sets, which will only be corrected as part of a full reprocessing procedure, coinciding with the production of the 2015 reference year. For example, an area of large (apparent) sealing increases in southern Spain is in reality caused by greenhouses in this area being included as sealed in 2009, but (by mistake) excluded in 2006.
    • Until the fully reprocessed time-series (2006-2009-2012-2015) becomes available in 2018, the existing status layers are only consistent with the change product for the 2006-2009 period, not for the 2009-2012 period. This is because the 2006-2009 product was produced and redelivered as a harmonised product and the 2006 errors found during the production of the change product were corrected in the 2006 status layer. This is not the case for the 2009-2012 product, where errors found in the 2009 product were not corrected in the 2009 status layer. This means that for some countries, an over-estimation of sealing in 2009 leads to (apparent) decreases in total area sealed between the two status layers that will only be corrected once the 2009 sealing is corrected. At the same time, most of these countries still show real increases in sealing as captured in the 2009-2012 change product. 
    • Where cloud gaps in 2009 were present, sealing values for 2006 were used in the 2009 product. This means that sealing increases for these areas is likely to be underestimated. This gap-filling will not be supported in 2015 and later updates (see section on 'Methodology for gap filling')
    • For imperviousness/soil sealing products, the term 'reference year' must currently be understood to comprise a maximum of 3 years (the year before the reference year, the actual reference year, and the year after the reference year for gap-filling). This means that the Earth observation data used to produce the '2009 reference year' are, in reality, a mix of 2008, 2009 and 2010 acquisitions. This situation is likely to improve (imagery is more likely to come from one single year) with the future availability of new sensors, such as Sentinel2.

Rationale uncertainty

No uncertainty has been specified


Other info

DPSIR: State
Typology: Descriptive indicator (Type A - What is happening to the environment and to humans?)
Indicator codes
  • LSI 002
Frequency of updates
Updates are scheduled every 3 years
EEA Contact Info info@eea.europa.eu