In this section, we present a collection of published definitions of land consumption, land take, soil sealing, and artificialization, highlight similarities and differences between the concepts in focus, and examine the interrelations between the concepts land consumption and land take (Section 3.1
). This is followed by information on relevant international monitoring and reporting systems and a comparison of selected conceptualizations and monitoring approaches at the national scale, highlighting differences between selected EU countries (Section 3.2
3.1. Definitions of Land Consumption, Land Take, and Related Concepts
From the studied scientific and policy relevant literature, as well as some internet sources, we extracted 30 definitions of the concepts of interest. Among those were eight definitions of land consumption, 13 definitions of land take, five definitions of soil sealing, and four definitions of artificialization (or of derivations of these terms, respectively; Table 1
). One of the definitions explicitly equates land consumption and land take (#1; this and all subsequent numbers marked with # refer to Table 1
). Another definition conceives land take as one of “three forms of land consumption” (#7).
Regarding the land use types they refer to, the extracted definitions range from broad to narrow (e.g., #1 vs. #2 for land consumption, or #16 vs. #9 for land take). Furthermore, the definitions vary in the degree to which they frame land take as being driven by urbanization or the growth of rural settlements (e.g., #2, #11), or by other specific developments like the building of infrastructures (e.g., #9 which was taken from a source dealing solely with the impacts of expanding transport networks). Many of the definitions of land consumption, land take, and soil sealing use the word “artificial” to specify the types of land, areas, or purposes under consideration (#5, #6, #8, #10, #12–#16, #20, #25, #26).
From the total set of collected definitions, one for land consumption (#1) encompasses the widest spectrum of land conversions and affected land types, including area used for (over-intensive) agriculture or forestry. In fact, according to this general definition of UN Habitat and the EEA, the term land consumption may be used to describe the spatial effects on land of any economic activity. In contrast, none of the collected land take definitions explicitly includes land conversions for the purpose of expanding agriculture or forestry. Rather, land take is described to convert (e.g., #10, #12) and diminish (#17) areas used for agriculture or forestry. Thus, if land consumption is conceived according to definition #1, all land take belongs to the category of land consumption, but not vice versa.
However, many of the reviewed sources conceive land consumption more narrowly than expressed in definition #1, i.e., as resulting primarily from urban developments (#2, #3, #5), and as referring only to processes that increase the area of “developed” or “artificial” land (#3–#6, #8). In these cases, the conceptual scopes of land consumption and land take are close, and possibly congruent; sometimes both concepts are used as synonyms (#11). In SDG indicator 11.3.1 (#3), the land consumption rate only refers to urban development [22
Land take is described by several of the collected definitions as a process that converts, and thereby diminishes, natural, semi-natural, forest, or agricultural land (#10, #12, #17–#19). Additionally, two land take definitions explicitly include inner-rural development and densification (#11, #16). In contrast, the densification of existing settlements, including the redevelopment of brownfields, is excluded from the definition of land consumption and land take according to the SDG indicator 11.3.1 (#3) and the EU indicator of land take (#10).
With regard to soil sealing and artificialization, the collected definitions suggest that both concepts are less equivocal than land consumption and land take. All five definitions collected for soil sealing describe this phenomenon as involving the covering of land, and four out of five mention that impermeable/impervious materials are used for this purpose (#22, #24–#26). One of the definitions for soil sealing (#22) include processes by which soils are not necessarily covered but changed in such a way that they lose the ability to absorb water (e.g., due to compaction).
The four definitions collected for artificial or artificialized areas/land surfaces (#27–#30) characterize the land they refer to as being in use for a range of different human activities; the exact set of activities varies among the definitions. The most encompassing definition of artificialized land refers to all human activities other than agriculture and forestry (#29).
Beside the differences in land use types to which the definitions refer, variation was observed with respect to the units of measurements the definitions suggest for the phenomenon in question. Applying this criterion, the following types of definitions emerge: (i) definitions from which an appropriate unit of measurement cannot be inferred, e.g., because they attribute the term in question to a process or activity rather than to its result (e.g., #18, #22–#25); (ii) definitions suggesting that the phenomenon in question should be expressed in absolute terms, i.e., as the total amount of area affected (e.g., #12, #27–#30); (iii) definitions suggesting that the phenomenon in question should be expressed in relative terms, i.e., as a percentage (affected area compared to a reference area, e.g., #3, #5), a rate (change in affected area over time, e.g., #11, #20), or in relation to the human population (e.g., change in affected area per capita, e.g., #4, #6). Measures of land use change that refer to the human population have been termed land use intensity [12
] or land use efficiency [29
3.2. Indicators for Monitoring Land Consumption and Land Take
As mentioned in the introduction, the SDG indicator 11.3.1 is composed of the quotient of the land consumption rate (LCR) and the population growth rate (also referred to as the land use efficiency [16
]). LCR is defined as “the percentage of current total urban land that was newly developed” [11
] and calculated as (ibid.):
= total areal extent of the urban agglomeration in km2
for past/initial year,
= total areal extent of the urban agglomeration in km2
for current year, and
= the number of years between the two measurement periods.
The methodology for computing the land consumption rate for the SDG indicator 11.3.1 is subject to a scientific debate [22
]. Promising advances have been made in using new global datasets, tools, and maps, specifically those subsumed under the header “Global Human Settlement Layer, GHSL” [2
]. However, there is no consensus yet on how to define the exact geographic boundaries of urban land and this poses a major challenge for assessing LCR [22
The land take indicator has been defined by the European Environment Agency to address “the change in the area of agricultural, forest and other semi-natural land taken for urban and other artificial land development. Land take includes areas sealed by construction and urban infrastructure, as well as urban green areas, and sport and leisure facilities” [14
] (see also definition #10 in Table 1
). It captures all transitions of “agricultural areas,” “forest areas,” “wetlands,” or “water bodies” into one of the following classes: “continuous urban fabric,” “discontinuous urban fabric,” “industrial or commercial units and public facilities,” “road and rail networks and associated land,” “port areas,” “airports,” “mineral extraction sites,” “dump sites,” “construction sites,” or “sport and leisure facilities” [14
According to this EEA methodology, land take may occur at any place where land classifies as natural, semi-natural, forest, or agricultural land, and these land type categories are almost exclusively assigned to land outside existing settlements. Urban green is seen as part of the settlement structure, and the respective land is regarded as “taken.” It follows that developments within the boundaries of a settlement are not counted as land take, even if they affect vegetated plots (e.g., brownfields, gardens). Inner-urban areas affected by densification thus change from a less intensely developed state to more intensely developed state, but as the area has already been counted as “urban,” the change has no effect on the land take statistics.
In contrast, successful land take compensation measures have an effect on the land take statistics because they change the status of areas from “taken” to “non-taken.” Such a “land return to non-artificial land categories” [14
] is also called re-cultivation, and can be achieved by unsealing or other conservation measures. By subtracting the re-cultivated area from the total area affected by land take, one arrives at the EEA indicator of net land take
(see definitions #13 and #14 in Table 1
). In the EU policy context, the concept of “net land take” plays a prominent role because it is the basis of the “no net land take until 2050” target. It is important to note that according to the EEA scheme, a distinction is made between “greening of urban areas” (which does not translate into a reduction of land take, because the area remains in the category “urban”) vs. “re-naturalisation of artificial areas” (which is a compensation measure and does reduce land take) [14
The EEA land take indicator is currently calculated using the CORINE Land Cover (CLC) Accounting Layers for the years 2000, 2006, 2012, and 2018. A serious limitation of CLC data is that small-scale developments cannot be detected in the Geographic Information System (GIS) vector dataset, as the Minimum Mapping Unit (MMU) for areal phenomena is 25 ha and the minimum width for linear phenomena is 100 m [48
]. This leads to the exclusion of most linear transport infrastructures from the EU land take assessment [14
] and may result not only in an underestimation of the total land take but also in flawed (counterfactual) judgements if, e.g., scattered sprawl (many smaller developments across the landscape) scores much better than coordinated, large developments. Due to its low resolution, CORINE based measures have proven inadequate for detecting relevant developments on a small scale [49
Furthermore, the CLC based method is not the only one by which land use is analysed across all EU Member States. The European Statistical Office (EUROSTAT), in close cooperation with the Directorate General for Agriculture and supported by the Joint Research Centre (JRC), samples the topsoil at more than 250,000 locations throughout the EU, approximately every three years. This survey is abbreviated LUCAS (which stands for Land Use/Cover Area frame statistical Survey) [50
], and the obtained data are used, inter alia, for monitoring the progress towards the SDG targets, because this is under the responsibility of EUROSTAT.
The differences in the methodologies between CLC based (obtained from remote sensing) and LUCAS based estimates of land use changes (obtained from in-situ sampling) have resulted in divergent conclusions about recent trends in land consumption/land take: While the EEA detected a reduction of land take in the last decade (from over 1000 km2
/year between 2000–2006 to 539 km2
/year between 2012–2018) [14
], EUROSTAT concluded that “the rate of land take has accelerated” between 2009–2015 [44
In addition to the two international schemes described above, several different national classifications and monitoring systems exist to assess and monitor land consumption and land take. These have their own specificities and are not necessarily harmonized and aligned to the international reporting schemes. The national approaches listed in Table 2
demonstrate how the conceptualization of land take differs between EU countries. For example, gardens are judged as being neutral to land take in Czechia but regarded as part of the settlement area in Austria, Belgium (Flanders), and Germany, thus contributing to land take. Areas for which building plans exist but that are not (yet) actually used for purposes related to settlements, infrastructure, or service provision do not count as being taken in Belgium (Flanders), but contribute statistically to land take in Germany.