LUCAS - Land use and land cover survey
From Statistics Explained
The European Union (EU) is composed of a diverse range of landscapes: it is home to a wide variety of flora and fauna and includes some of the most and least densely populated areas of the world. This background article provides information on the Land Use/Cover Area frame Survey (LUCAS), a survey that provides harmonised and comparable statistics on land use and land cover across the whole of the EU’s territory - a toal area of just under 4.5 million square kilometres (km²) for the EU-28.
The data collected by LUCAS provides harmonised information for studying a range of socioenvironmental challenges, such as land take, soil degradation or biodiversity.
- 1 Defining land use, land cover and landscape
- 2 The LUCAS survey
- 3 LUCAS — the historical perspective
- 4 How is LUCAS conducted?
- 5 What sort of information is compiled?
- 6 How is quality assured?
- 7 What results are made available?
- 8 How can LUCAS data be used?
- 9 See also
- 10 Further Eurostat information
- 11 External links
Defining land use, land cover and landscape
Land cover refers to the bio-physical coverage of land (for example, natural areas, forests, buildings and roads or lakes).
Land use refers to the socioeconomic use that is made of land (for example, agriculture, commerce, residential use or recreation); at any one place, there may be multiple and alternate land uses.
Landscape refers to an area of land whose character and functions are defined by the complex and regionally-specific interaction of natural process with human activities that are driven by economic, social and environmental forces and values. Landscape can be analysed by taking account of elements such as landscape diversity, the importance of linear features, and the degree of landscape fragmentation.
The LUCAS survey
The Land Use/Cover Area frame Survey (LUCAS) is a harmonised in situ land cover and land use data collection exercise that extends over the whole of the EU’s territory. An in situ survey implies that data are gathered through direct observations made by surveyors on the ground.
LUCAS is based on statistical calculations that interpret observations in the field. It is based on a standardised survey methodology in terms of a sampling plan, classifications, data collection processes and statistical estimators that are used to obtain harmonised and unbiased estimates of land use and land cover.
LUCAS — the historical perspective
LUCAS was initially developed to provide early crop estimates for the European Commission. The survey started as a pilot survey across a limited number of EU Member States; the first survey was held in 2001.
Over time, the survey has become a key tool for policymakers and statisticians alike, with increasing amounts of data on different forms of land use and land cover in the EU. In 2006, the sampling methodology changed and its focus shifted from an agricultural land survey to a broader land cover, land use and landscape survey. In the same year, a three-yearly interval was introduced as the frequency for carrying out the survey.
The 2009 survey saw a marked expansion in terms of the geographical coverage of LUCAS with results made available for 23 of the then EU-27 Member States (Bulgaria, Cyprus, Malta and Romania were not covered).
The classifications used for categorising data on land use and land cover were adapted in order to be more closely aligned with international standards.
The latest survey was conducted in 2012 and covered all of the then EU-27 Member States.
How is LUCAS conducted?
LUCAS is a two phase sample survey. The LUCAS first phase sample is a systematic sample with points spaced 2 km apart in the four cardinal directions covering the whole of the EU’s territory; it therefore includes around 1.1 million different points. Each point of the first phase sample is photo-interpreted and assigned to one of the following seven pre-defined land cover strata — arable land, permanent crops, grassland, wooded areas and shrubland, bareland, artificial land, and water.
From the stratified first phase sample, a second phase sample of points, the field sample, is drawn. The stratified sample is selected independently in each NUTS level 2 region fixing precision targets for the estimates of the main land cover classes. Each of these field samples is classified during the field visit according to the full classifications of land use and land cover. For more information on the methodology, follow this link.
A panel approach assures that a certain percentage of points are surveyed in successive LUCAS campaigns. Points above 1 500 metres and far from the road network are considered inaccessible and excluded from the sample of points to be visited to limit the cost of the data collection exercise.
The sample design is taken into account for the computation of the final estimates by calculating appropriate weights for each surveyed point. The final weights are computed taking into account two elements:
- i) the weight of the first phase stratification assigned to each of the seven strata (in each NUTS level 2 region) and;
- ii) the second phase sampling weight resulting from the combination of the sampling weight needed to extrapolate what is observed in the sample to the total population, and the weight for missing data adjustment to account for points missing 'by design' (those points excluded because they are inaccessible).
Since the same variable (land cover) was used to stratify the points in the first phase and to classify them in the second phase, the observed transitions from one strata to another are over-emphasised when they occur among a prevalent strata (for example, woodland) when compared with a less common strata (for example, artificial land). Points whose cover changes from woodland to artificial land keep the first phase weight of the original stratum (first phase weight for woodland = 0.5694030) which is larger than the first phase weight for other points with artificial cover (first phase weight for artificial land = 0.04962427).
A regular update of first phase stratification results needs to be performed in order to reduce to a minimum the potential bias introduced. As a consequence a comparison of LUCAS data over the years should generally be avoided, especially when a lengthy period of time has elapsed since the first phase sample was stratified.
The 2012 survey was based on 270 000 points/observations, which were visited by 750 field surveyors (mostly agricultural and forestry engineers); the field survey took place between March and September 2012.
The use of an area frame survey reduces the statistical burden on farmers and other land owners as they do not need to respond to questionnaires.
LUCAS surveyors go to the points and observe the land cover, land use and environmental parameters they find on the ground. The surveyor documents the land cover and land use according to harmonised classifications. The concept of ‘land’ is extended to inland water areas (such as lakes, rivers, estuaries or lagoons). The surveyor also collects information relating to the percentage of land cover within a specific window of observation, the area size, the width of any specific features, the height of any trees, as well as information on land and water management (for example, grazing or irrigation). Surveyors receive training before going into the field: they have a set of supporting documents, instructions on how to carry out the survey, and a set of quality control procedures. As such, considerable efforts are made to ensure that each of the surveyors applies the same methods when visiting the assigned geographical point. They fill in a questionnaire with a series of land cover and land use parameters. They also take a series of photographs at each point, of the point itself, as well as pictures in all four cardinal directions (north, south, east and west), before walking 250 metres in an eastwards direction (a ‘transect’), recording all of the different land cover and linear elements. These linear features include elements such as walls, hedges, roads, railway lines, irrigation channels or electric power lines.
The information collected for this transect can be used to analyse the fragmentation, richness and diversity of landscapes.
In addition, specific modules can be included in the survey, such as, for example, a soil survey. In the 2009 LUCAS data collection exercise a top soil sample was taken for 10 % of the surveyed points; other specific modules might be included in the future.
Some examples of decisions taken in the field:
|Example A||the point is located in a common wheat field — the land cover is cropland of common wheat and the land use is agriculture|
|Example B||the point is located on the lawn of a campsite with trees — the land cover is grassland with sparse trees/shrub cover and the land use is a holiday camp|
|Example C||the point is located on a road that is wider than three metres — the land cover is a non-built-up linear feature and the land use is road transport|
|Example D||the point is located in an urban park with mainly broadleaved trees — the land cover is broad-leaved woodland and the land use is leisure; for this particular example the forest species would also be noted by the surveyor (for example, beech)|
|Example E||the point is located in the parking area of a supermarket — the land cover is a non-built-up area feature and the land use is commerce, finance, business|
LUCAS surveyors are sometimes confronted with unusual, even perilous situations, as some points may prove extremely difficult to reach, for example when trying to access flooded areas, encountering rock faces or disturbing (wild) animals. The safety of surveyors is clearly paramount and in some cases the specified point cannot be surveyed.
What sort of information is compiled?
The classifications used within LUCAS are comparable with other statistical standards, for example, standards used in the EU’s farm structure survey (FSS), those used by the United Nations’ Food and Agriculture Organisation (FAO), or the European Nature Information System (EUNIS) for classifying forestry types and areas.
Land cover is classified according to the following classification (only the first two levels of the hierarchy are shown — there are a total of 76 subclasses):
Land use is classified according to the following classification (there are 14 main categories and a total of 33 distinct classes):
The analysis of the soil samples from the LUCAS 2009 data collection exercise provided information on:
- soil types;
- soil textures;
- pH levels;
- organic carbon;
- phosphorous, nitrogen and extractable potassium;
- soil erosion;
- susceptibility to compaction.
For more information: background to data compilation
How is quality assured?
The results collected by surveyors are subject to a detailed quality check. An automated quality check, checking for completeness and consistency, is carried out during the compilation phase or when the data collected in the field is uploaded to a central data repository. A second level of quality controls is carried out at the regional or central offices, where all of the surveyed points are checked visually. A further quality control is done by an independent quality controller and includes:
- interactive control of accuracy and compliance to the quality requirements as defined in the LUCAS framework for 36 % of the points;
- the first 20 % of points assigned to a surveyor are controlled in their entirety to detect early any systematic errors.
All the available information is used for carrying out the controls, including:
- the ground documents that the surveyors receive to locate the point;
- the actual photos as well as historic photos and data in case the point has been surveyed previously;
- the GPS tracks that show the complete path that has been followed by the surveyor;
- any comments made by surveyors and local controllers, which are transmitted together with the data.
Those data points requiring correction or additional clarification (around 7 % of the total number of points) are rejected and sent back to the field contractors for correction.
What results are made available?
There are two main types of information presented to users: aggregated statistical tables and elementary data (linked to individual points).
The aggregated results for the EU-27 and national totals show the land cover and land use for each of the main categories and their subclasses (see below for more details of the classifications used), as well as woodland areas by size and by canopy cover. The statistics can also be analysed at a more detailed level according to the NUTS classification — the EU’s classification of regions — providing data for the almost 300 NUTS level 2 regions within the EU-27. The statistical tables are supplemented by indices relating to landscape composition, richness, structure, dissection and diversity. These aggregated data are available here.
The elementary data in its most disaggregated form (in situ micro-data) can be accessed for individual points. The data are presented in a tabular format in country-specific files, available here.
Since the 2006 reference period, Eurostat has also made a photo archive from LUCAS available. Photos can be requested by using the online form, available here.
Furthermore, there is a LUCAS online viewer for the 2009 and 2012 data collections, available here.
LUCAS soil data is available here.
An overview of the aggregated statistical results from LUCAS can be found in the following articles:
How can LUCAS data be used?
Data on land cover and land use can be used for a variety of environmental and socioeconomic projects linked to a range of policy areas:
- the European Commission’s Directorate-General for Agriculture and Rural Development, for evaluating the impact of agriculture on the environment through agri-environmental indicators (AEI), including for organic soil matter and soil erosion as well as indicators on the degree of artificiality and the physical structure of landscapes within the framework of the integration of environmental concerns into the Common Agricultural Policy (CAP) post-2013;
- the European Commission’s Directorate-General for the Environment, for resource efficiency indicators that form part of the Europe 2020 strategy and for soil protection;
- the European Commission’s Directorate-General for Climate Action for analysing climate change as part of the European climate change programme;
- the European Commission’s Directorate-General for Enterprise and Industry, where in situ data from LUCAS contribute to the production, verification and validation processes relating to pan-European data sets describing the main land cover types, which are derived from satellite images, as conducted by Copernicus observation programme;
- the European Environment Agency, for their core set of indicators (CSI), climate change indicators (CLIM) and streamlining European biodiversity indicators (SEBI);
- the European Commission’s Directorate-General on Climate Action, for land use, land use change and forestry (LULUCF) statistics in relation to the reduction of greenhouse gas emissions.
Specific examples on how the data is used within the European Commission can be found here.
LUCAS data also provides a rich source of information for the research community.
- Land cover statistics
- Land cover, land use and landscape
- Land cover and land use statistics at regional_level
- Landscape structure indicators from LUCAS
Further Eurostat information
- Diversified landscape structure in the EU Member States
- New insight into land cover and land use in Europe
- Land cover (lan_lcv)
- Land cover overview (lan_lcv_ovw)
- Land covered by artificial (lan_lcv_art)
Methodology / Metadata
- Land cover and land use, landscape (LUCAS) (ESMS metadata file — lan_esms)