The LUCAS 2006 project – A new methodology - Institute for [PDF]

2001 - 2003. 2006. State. Nbr of PSUs. Nbr of SSUs. Nbr of points Area (km²). Austria. 255. 2528. 4859. 83891.65. Belgi

0 downloads 3 Views 686KB Size

Recommend Stories


Project Methodology
Live as if you were to die tomorrow. Learn as if you were to live forever. Mahatma Gandhi

[PDF] New The Happiness Project New Book
Goodbyes are only for those who love with their eyes. Because for those who love with heart and soul

Methodology for New Business Development
Learn to light a candle in the darkest moments of someone’s life. Be the light that helps others see; i

A new proof of Lucas' Theorem
You miss 100% of the shots you don’t take. Wayne Gretzky

Starting a New Project
If you want to go quickly, go alone. If you want to go far, go together. African proverb

Traditional Project Management Methodology
How wonderful it is that nobody need wait a single moment before starting to improve the world. Anne

2006 Transmission Loss Factor Methodology
The only limits you see are the ones you impose on yourself. Dr. Wayne Dyer

America's Best Hospitals 2006 Methodology
When you do things from your soul, you feel a river moving in you, a joy. Rumi

Download Project Profile 2003-2006 (PDF: 7MB)
If you want to become full, let yourself be empty. Lao Tzu

New Generalizations of Lucas Numbers
You have survived, EVERY SINGLE bad day so far. Anonymous

Idea Transcript


The LUCAS 2006 project – A new methodology Pascal JACQUES*, Javier GALLEGO** *Eurostat – Unit E1: Agriculture Statistics - Methodology **JRC-ISPRA – MIPSC/AGRIFISH unit Summary: LUCAS is a pilot project launched by Eurostat in close co-operation with the Directorate General of Agriculture with the technical support of JRC Ispra. LUCAS is an area frame statistical survey aiming to obtain harmonised data at EU level on land use, land cover and environment. The survey has been carried out on EU-15 in 2001 and 2003 and a new survey is planned at EU level for 2006. The present paper presents an overview of the new methodology, the reasons for its new design and some preliminary results obtained during a pre-pilot operation in 2005. Keywords: area frame sampling, land use, land cover, stratification, agricultural statistics.

1.

INTRODUCTION

Land cover and land use receive an increasing importance in the definition and evaluation of sectorial common policies (integration of environment or sustainable development into agricultural, regional or transport policies) as well as in the daily local management of the territory. Although similar, the concepts of land use and land cover provide different representations of the situation on the ground. Land cover deals with the crop or artificial coverage at ground level (building, maize growing, coastal lagoon) whereas land use relates to the function or destination of the output (industrial, commercial or residential building, green or grain maize, lagoon used for leisure purposes, as a nature reserve or for fish farming). If statistics on land use and land cover at EU level are based for the major part on the own Member States' statistics, this approach has some limitations in terms of comparability and harmonization of produced information. In 1998, the Future Agricultural Data Outline seminar (FADO) 1 concluded to the need to realize EU level surveys when they are necessary to obtain important data for the management of the Common Agricultural Policy (CAP) and when they cannot be obtained within the required deadlines and with sufficient quality when aggregating national data. Agricultural areas were mentioned as potential data to be obtained by that mean. This issue has been reemphasized during the 26th European Advisory Committee on Statistical Information in the Economic and Social Sphere (CEIES) 2 seminar in September 2004 also requesting for a more rapid delivery of survey results.

1

Acts of FADO Seminar – Future Agricultural Data Outline – 13-15 May 1998. Acts of the 26th CEIES Seminar – European agricultural statistics – Europe first or Europe only ? – 9-10 September 2004. 2

2.

CONTEXT

In order to experiment the integration of land use and land cover data at European level through harmonisation of nomenclatures and collection methods, the LUCAS “Land Use/Cover Area frame statistical Survey" pilot project was launched by Eurostat in 2001, in close collaboration with the Directorate General for Agriculture and with the technical support of the Joint Research Centre ISPRA. Area frame surveys are typical approaches to gather land cover and land use data. In contrast to mapping approaches (e.g. the CORINE Land Cover project), area frame sampling is a statistical method. Based on the visual observation of sample georeferenced points, area estimates are computed and used as a valid generalisation without studying the entire area under investigation. The approach has also the important advantage of not involving or disturbing the land owners and the farmers. The project has been implemented following the Decision N°1445/2000/EC of the European Parliament and of the Council of the 22.05.2000 “On the application of areaframe survey and remote-sensing techniques to the agricultural statistics for 1999 to 2003”, continued until 2007 by Decision 2066/2003/EC of 10 November 2003 and extended to EU-N10 by Decision 786/2004/EC of 21 April 2004. LUCAS has been implemented in 2001 in 13 EU Member states. Due to the Foot and Mouth Disease, the LUCAS survey was postponed in 2002 in United Kingdom and in Ireland at the same time as in Estonia, Hungary and Slovenia. The survey was carried out again in 2003 in all EU Member States (15) plus Hungary, allowing improvement of the data collection system and analyses of land use and land cover changes (2001-2003). Results have largely been published on the two surveys and lessons have been drawn in order to prepare the next survey forecasted in 2006 on the 23 Member States 3 . The objective of the present paper is to present the new methodology developed for the survey in 2006, based on the conclusions of LUCAS 2001 and 2003 surveys and on the experiments of JRC Ispra in Greece in 2004. 3.

PURPOSE OF THE PROJECT

The overall objectives of the LUCAS project are: •

To carry out an area-frame survey based on points in order to collect information on land cover and land use, within an acceptable period, particularly in terms of its agricultural component in the broad sense;



To develop a standard survey methodology in terms of the sampling plan, the nomenclature and the collection and use of data to obtain harmonised data (unbiased estimates) at EU level of the main Land Use / Cover areas and changes limiting the workload placed on farmers. Precision is expected to be around or better than 2% 4 for main categories like wheat, cereals, arable land, permanent grassland, permanent crops, forests, urban areas, inland waters.

3

Malta and Cyprus are not included. To allow comparison with the results provided by Council Regulations (EEC) N° 959/93 on other crop statistics and (EEC) N° 837/90 on cereals production

4



To offer a common sampling base (frame, nomenclature, data treatment) that interested Member states can use to obtain representative data at national or regional levels by increase of the sampling rate, respecting the general LUCAS approach.



To extend the scope of the survey from the normal agricultural domain to aspects relating to the environment, multi-use, landscape and sustainable development.

4.

THE NEW LUCAS METHODOLOGICAL IMPLEMENTATION

4.1.

Definition of the grid

Apart facilities in the management of the data, the insensitivity to country borders and the ease of extension of the system, many reasons justify the selection of a common grid covering the 25 Member States including the possibility to allow a better harmonisation and exchange of cross-sectorial information and provide a cognitive added value. A new base grid of 1 Km size, making it an easy multiple of other already existing grids, has therefore been generated following the INSPIRE’ recommendations 5 . The projection used is the Lambert Azimuthal Equal Area coordinate reference system. The grid is squared, with origin: 4,321,000 m west of centre point of the projection (52N, 10 E), and 3,210,000 m south of projection centre point (52N 10E) and orientation: south – north, west – east. Each point has been given a unique numeric code going sequentially from south-west to north-east. 4.2.

Sample design

The unstratified two-stage sampling design used in 2001 and 2003 has been substituted by a two-phase sampling of unclustered points with stratification after the first phase. The two-phase sampling chosen for the 2006 survey allows straightforward calculation of the estimates, improves the efficiency of the stratification, reduces the variance by an estimated factor of three, avoids the problems of not complete PSUs 6 and is more logical in the implementation perspective. Moreover, first experiences from 2005 surveys and preparation of 2006 survey show that the single level design seems to also allow reducing the survey costs. The systematic sample (base sample) linked to the 1 Km grid defined previously corresponds to around 4,000,000 points for entire European Union. The LUCAS master sample is a subset of the base sample corresponding to a 2 Km grid created by using all the even points of the base sample and therefore consists of around 1,000,000 points. Points located on small islands were excluded from the sample (examples of excluded islands are: Baleares, Azores, Canary islands, Cyprus, Malta, the Greek islands except Creta).

5

1st Workshop on European Reference Grids held in Ispra, 27-29 October 2003 PSU : Primary Sampling Unit – In LUCAS 2001 methodology, PSUs are defined as cells of a regular grid with a size of 18 x 18 km.

6

The master sample has been intersected with NUTS boundaries (source: GISCO 7 ) to extract points on the EU territory and allocate them by country, as well as with a digital terrain model to discards points above 1200m altitude and with the Corine Land Cover (CLC 2000) dataset to provide supplementary information to help the photointerpretation process. 4.3.

Stratification

Each point of the master sample is photo-interpreted in order to classify the sample into seven strata (“arable land”, “permanent crops”, “permanent grassland”, “wooded areas, shrubland”, “Bare land, low or rare vegetation”, “water” and “artificial land”). This photo-interpretation is based on the most recent ortho-photos or, where ortho-photos are not available, on satellite imagery (Corine Image 2000 Landsat Images).

Figure 1 - Results of the stratification process

4.4.

Sampling

From the stratified master sample, a sub-sample of points (field sample) is extracted to be classified by field visit according to the full land nomenclature 8 . Around 250.000 points are necessary in the field sample to reach the desired precision. 7

Geographic Information System of the European Commission - EUROSTAT Three Levels nomenclature for Land Cover – 54 classes and three levels nomenclature for LandUse – 33 classes. http://forum.europa.eu.int/irc/dsis/landstat/info/data/Documentation.htm

8

To arrive to the requested precision in optimising the available budget, different sampling rates have been applied in the different strata focussing on the strata of interest (Table 1). Stratum Arable Land Permanent Crops Grassland Woodland, Shrubland Bare Land Artificial Areas Water

Sampling rate 50% 50% 40% 10% 10% 10% 10%

Table 1 – Sampling rates applied by stratum

Table 2 presents the sample repartition by country used in 2001/2003 and the calculated sample size to be used in 2006. State Austria Belgium Germany Denmark Spain Finland France Greece Ireland Italy Luxembourg Netherlands Portugal Sweden United Kingdom EU15 Slovenia Estonia Latvia Lithuania Czech Republic Hungary Poland Slovak Republic EU23

2001 - 2003 2006 Nbr of PSUs Nbr of SSUs Nbr of points Area (km²) 255 2528 4859 83891.65 100 989 2292 30558.92 1105 10981 26452 357296.19 147 1373 3814 42970.44 1268 12670 33607 506711.69 1073 10410 10664 336411.75 1702 16916 40380 549101.14 419 4051 7998 131758.89 218 2163 5631 69993.41 941 9275 20247 301617.04 8 80 229 2584.52 117 1154 2870 37198.13 277 2731 4995 91857.19 1407 13808 14810 446691.84 775 7499 19236 244168.09 9812 96.633 198084 3232810.9 1051 49030 2346 45227 4061 64589 5139 65300 5420 78865 8133 93030 23076 312685 3230 20273 264829 3.961.809.89

Table 2 - Samples used for the survey in 2001/2003 and the forecasted 2006 survey

4.5.

Sample drawing

Stratified systematic sampling rates in different strata may be carried out with independent grids of different steps in each stratum. The approach leads unfortunately to samples in which the spatial distribution is not optimal as points sampled in different strata can be close to each other and give some redundant information, as spatial correlation happens also between strata.

A more homogeneous layout of the sample has been obtained by dividing the systematic grid into subgrids or replicates and selecting a variable number of replicates for each stratum. To this purpose, the basic sampling grid has been divided into squares (9 by 9 that is 18 Km by 18 Km) blocks of 81 points each. The set of points with the same relative position in the block is named a replicate. If the order of replicates is selected at random, it may happen again that selected replicates are close to each other and large areas remain still un-sampled. To solve this problem, the following procedure is applied: The first replicate is selected at random; To avoid that the second replicate sampled is too close to the first, a distance restriction is imposed in the random sampling, forbidding replicates below a certain distance, or selecting the replicate at the maximum possible distance of the first replicate. In figure 2, for the 9 by 9 block size selected, assuming that the shaded points belong to the same domain (NUTS 2 region and stratum), replicate 1 consists of the highlighted points.

Row in block ↑

9

23

66

44

68

10

48

16

42

76

23

66

44

68

10

48

16

42

76

8

71

12

69

25

51

29

60

37

7

71

12

69

25

51

29

60

37

7

7

20

79

18

72

78

3

31

63

70

20

79

18

72

78

3

31

63

70

6

33

1

80

11

59

32

38

9

64

33

1

80

11

59

32

38

9

64

5

28

40

26

49

55

17

53

50

77

28

40

26

49

55

17

53

50

77

4

45

27

41

67

6

65

15

73

5

45

27

41

67

6

65

15

73

5

3

35

39

13

36

62

21

57

24

47

35

39

13

36

62

21

57

24

47

2

8

58

74

46

14

75

2

56

34

8

58

74

46

14

75

2

56

34

1

43

30

4

54

61

19

81

22

52

43

30

4

54

61

19

81

22

52

9

23

66

44

68

10

48

16

42

76

23

66

44

68

10

48

16

42

76

8

71

12

69

25

51

29

60

37

7

71

12

69

25

51

29

60

37

7

7

20

79

18

72

78

3

31

63

70

20

79

18

72

78

3

31

63

70

6

33

1

80

11

59

32

38

9

64

33

1

80

11

59

32

38

9

64

5

28

40

26

49

55

17

53

50

77

28

40

26

49

55

17

53

50

77

4

45

27

41

67

6

65

15

73

5

45

27

41

67

6

65

15

73

5

3

35

39

13

36

62

21

57

24

47

35

39

13

36

62

21

57

24

47

2

8

58

74

46

14

75

2

56

34

8

58

74

46

14

75

2

56

34

1

43

30

4

54

61

19

81

22

52

43

30

4

54

61

19

81

22

52

1

2

3

4

5

6

7

8

9

1

2

3

4

5

6

7

8

9

Column in block →

Figure 2 – Examples of four 9*9 blocks with replicates order

Replicates are then selected successively (starting with replicate 1) until the required sample size by domain is reached. From the replicate with the highest number, points are randomly selected.

4.6.

Overview of the process

The entire 2006 survey process could therefore be split and summarized into 4 steps: • Selection of the master sample. The LUCAS master sample is the intersection points of a 2 km-grid covering the territory of the EU, consisting of 990,816 georeferenced points. • Photo-interpretation and stratification. Each point of the master sample is photointerpreted, and classified into 7 strata. • Selection of the field sample. From the stratified master sample, a sub-sample of points (around 250.000 points for 2006) is extracted in order to be classified by field visit according the full land nomenclature. • Field work. In the period of March – June 2006, surveyors will collect the land information on the field sample in order to provide first early estimates of "southern" countries by mid-June and on the Union by mid-July.

Grid 2 km

Photo-Interpretation 1,000,000 points Stratification

Orthophotos

Survey

Survey forms

Sampling 250,000 points

Compute estimates LC/LU results

Figure 3 – Overview of the entire survey process

5.

ESTIMATION OF THE MEAN SURFACE AND ITS ACCURACY

The mean of the surface of a specific crop c can be estimated according to the usual formula for stratified random sampling (The notation used is the notation in Cochran (1977): capital letters refer to characteristics of the population, and lowercase letters to those of the sample): nh L

L

y st = ∑ wh yh = ∑ wh

∑ yhi

i =1

where wh =

nh h =1 h =1 D is the surface of the area of interest Dh is the surface of stratum h.

Dh D

For the calculation of variance of land cover area : ⎛ 1 1⎞ N −n 2 − ⎟⎟ + v( y dst ) = ∑ wh sh2 ⎜⎜ wh ( y h − y dst ) (Cochran 12.24) ∑ h ⎝ n′ν h N ⎠ ( N − 1)n′ h where N is the size of the population in the whole region.

Since a point is considered to have a size 3x3 m N = region area in hectares / 0.0009 (1/N is very small and can be disregarded)

sh2 is an estimate of the variance of y (not of y ) In this case we use an estimate of the local variance: 2 ∑ δ ij yi − y j sh2

= (1 − f h )

(

i≠ j

)

2 ∑ δ ij i≠ j

Where

fh =

nh

Nh

1 d (i, j ) For countries where a stratification based on photo-interpretation of a pre-sample of points (regular grid) is applied (2006 survey, Greece 2004, Lithuania and Latvia 2005), we have, for each i: ⎧⎪ 1 if j is among the 8 closest points to i in the stratum δ ij = ⎨ d (i, j ) ⎪⎩ 0 otherwise When there is no stratification, like in Poland 2005, an alternative is: ⎧⎪ 1 if d (i, j ) < 3000 m δ ij = ⎨ d (i, j ) ⎪⎩ 0 otherwise

δ ij

6.

is a decreasing function of the distance between i and j :

δ ij =

PRELIMINARY RESULTS

The results of the Greek survey made by JRC Ispra in 2004 confirmed the expected improvement in precision with the adoption of the modified point frame methodology. It served as a reference for drafting the specifications for the LUCAS 2005 and 2006 surveys.

The pre-pilot survey made in 2005 and the answers to the call for proposals for field work in 2006 confirmed that:



The new methodology is applicable and experience has been gained in 2005



It is possible to provide early estimates, even for northern countries like Lithuania and Latvia, on the 15th of July.



The analysis of the costs for ground survey shows that costs per points are more or less similar between the 2001/2003 and the 2005/2006 surveys. Some costs in very large countries like FI and SE in 2006 were due to the important split of the survey points and were not taken into account for the cost comparison. PL342 - Lomzynski Area (Km²) LUCAS A - ARTIFICIAL LAND A1 - BUILT-UP AREAS A2 - ARTIFICIAL NON BUILT-UP AREAS

Census 2002

131.43 59.74

PL622 - Olsztynski

LUCAS

Census 2002

152.71 104.49

71.69

LV - LATVIJA New Cronos LUCAS 2005 1316.05 721.30

48.22

1923.67 1712.58 55.76 123.47 458.02 238.96 207.10 629.27 159.31 131.43 27.88 0.00

2140.00 1832.70 158.8 64.3 486.5 154.4 125.9 842.8 286.10 247.5 34.1 4.5

2616.15 2113.82 996.63 180.84 305.42 152.71 96.45 381.77 124.58 60.28 36.17 28.13

2307.80 1900.00 852.3 204.4 222.8 107.7 72.8 440.0 115.30 80.2 31.8 3.3

6104.06 4486.62 1964.23 1526.67 328.56 460.82 18.02 188.32 683.74 507.42 96.18 80.15

3.98

2.90

277.29

266.80

7.97

10.20

60.28

39.83 11.95

8.10 8.1

40.19 4.02

C - WOODLAND C1 - FOREST AREA C11 - Broadleaved forest C12 - Coniferous forest C13 - Mixed forest C2 - OTHER WOODED AREA

1286.43 1230.67 231.00 872.22 127.45

1373.8

55.76

D - SHRUBLAND

63.72

E - GRASSLAND

1804.18

F – BARE LAND

59.74

G - WATER

1402.5

1807.71 714.73

594.75

B - CROPLAND B1 - CEREALS B11 - Common wheat B13 - Barley B14 - Rye B15 - Oats B16 - Maize B18 - Other cereals B2 - ROOT CROPS B21 - Potatoes B22 - Sugar beet B23 - Other root crops B3 - NON PERMANENT INDUSTRIAL CROPS B4 - DRY PULSES, VEGETABLES AND FLOWERS B7 - PERMANENT CROPS B71 - Apple fruit

LT - LIETUVA New Cronos LUCAS 2005

1092.98 4750.00 4750 1890 1510 400 700

13587.67 10220.72 3703.33 4314.39 789.76 537.83 102.96 772.45 904.33 572.46 205.99 125.88

11713.03 9514.98 3680.84 3495.04 485.23 828.92 153.93 28.645 842.03 632.35 209.68

629.08

1506.47

1095.27

16.30

124.36

394.82

260.75

9.40 9.4

180.25 131.78

561.32 298.12

3423.91 3383.72 486.26 1977.19 920.28

3504.8 31462.61 30375.28 5333.65 8877.39 16164.23

19947.21 19397.08 4376.09 7407.38 7613.60

40.19

1087.33

550.13

100 460

345.61

42.4

2571.00

2408.99

3243.07

1599.3

20111.8

23345.58

969.06

1853.23

124.58

39.83 425.98 2054.39 Table 3 - Results of the survey in Poland, Lithuania and Latvia in 2005

2269.50

7.

REASONS FOR CHANGING THE METHODOLOGY - LESSONS FROM THE 2001 AND 2003 SURVEYS

A certain number of statistical quality problems have appeared in the implementation of the surveys in 2001 and 2003 which needed corrections through the redesign and the refinement of the survey methodology. 7.1.

Accuracy of estimates

Before the project was implemented, it had been estimated that the accuracy of the sample design would allow for about 2% for the main classes of land use and land cover at EU level. Those objectives were setup in order to permit comparison with the accuracy of the data transmitted by the Member states in the framework of the Council Regulations (EEC) No 959/93 on other crop statistics and the 837/90 on cereals production. The obtained accuracy is sufficient for important land covers without reaching totally the objectives of a precision lower than 2%. At the finest level of the land cover nomenclature, the mentioned accuracy is obtained for areas above 37 Mha (Table 4). LUCAS results present strong differences from the National data (comparison performed mainly for the third level of the nomenclature) in those cases where:



The surface of a specific land cover is small (the coefficient of variation is large)



The definition of the matched categories does not fully coincide (other fresh vegetables, other cereals, permanent grasslands, fallow land)

Of course, the results should be analysed taking into account the pilot approach of the survey and the limited size of the sample. In order to reach the precisions mentioned above, the sample size should be increased also to cover the EU-N10 Member States and the sample methodology improved to reduce the survey cost and simplify the survey procedures.

LAND COVER CATEGORY

Estimated Area (1000 ha)

% of total EU15 area

CV

ARTIFICIAL LAND

16.070

4,96

2,16

CROPLAND

83.335

25,72

0,95

CEREALS

41.395

12,78

1,32

Common wheat

13.106

4,04

2,25

4.319

1,33

4,73

10.708

3,30

2,50

Rye

1.222

0,38

7,65

Oats

2.265

0,70

5,73

Maize

8.529

2,63

2,71

284

0,09

21,16

Durum Wheat Barley

Rice Other cereals

962

0,30

8,52

ROOT CROPS

2.785

0,86

4,81

NON PERMANENT INDUSTRIAL CROPS

6.095

1,88

3,59

PERMANENT CROPS: FRUIT TREES, BERRIES

3.812

1,18

5,14

8.510

2,63

3,44

112.247

34,64

0,80

OTHER PERMANENT CROPS WOODLAND

SHRUBLAND

28.159

8,69

1,99

PERMANENT GRASSLAND (incl. pastures)

52.291

16,14

1,16

8.898

2,75

3,72

23.016

7,10

2,07

BARE LAND WATER AND WETLAND EU15 TOTAL AREA

324.016 Table 4- Accuracy of land cover results in 2003

7.2.

Comparability

7.2.1.

Land cover/land use

The added-value of the project is the possibility to compare the observations done in two successive surveys in order to detect differences and compute change matrices. In the change matrices, all flows between the land cover/land use classes become visible, i.e. flows of land cover/land use data reflecting the development of land such as urban sprawl, extension of forest areas etc. Through the LUCAS surveys, some land cover/land use flows could be exemplified, such as:



Abandonment of agricultural land



Intensification of agricultural production



Uptake of agricultural land by different economic sectors (IRENA 9 indicator N° 12, “Topological Change”)



Modification of agricultural land (IRENA indicator N° 24, “Land use change”)



Aforestation, deforestation, etc.

LUCAS data may also be used to report about detailed changes in land cover/land use and how these physical changes in terms of land consumption can be further interpreted. However, due to the short time period between the two surveys, the main changes occur in agricultural land as part of the crop rotation systems applied by farmers. Longer time periods are necessary. Moreover, an increased sample could also substantially improve the added value of the survey in enabling the analysis at a finer geographical breakdown and in allowing a further disaggregation of the nomenclature. Table 5 displays not only the “regular” changes but others less realistic changes mainly based on wrong interpretation of survey rules or location errors. The quality of the survey plays an important role in the set up of the matrices and choices made for the 2003 survey (in a double-blind mode) have increased the errors due to bad localisation of the point during the revisit of the point from 2001 to 2003. The survey procedures should therefore allow for more precise information gathering on the location of the point for further visits. To allow this, the use of GPS and photos of the point location have been requested in 2006 and will be provided to surveyors in the post-2006 surveys.

9

COM 2000 (20) “Indicators for the Integration of Environmental Concerns into the Common Agricultural Policy” and COM(2001) 144

Land Cover

2001 Survey

Share of total land

2003 Survey

Share of total land

Net change Net change (2001-’03) %

Built-up areas

1 405

1.5

1 642

1.7

237

16.9

Non built-up areas

1 151

1.2

1 327

1.4

176

15.3

traffic infrastructure

1 914

2.0

2 008

2.1

94

4.9

artificial green areas

2 763

2.9

2 649

2.8

- 114

-4.1

15 490

16.4

15 440

16.3

-50

-0.3

3 472

3.7

3 570

3.8

98

2.8

15 362

16.3

15 433

16.3

71

0.5

fallow land

2 273

2.4

1 864

2.0

- 409

-18.0

mixed agriculture

2 133

2.3

2 720

2.9

587

27.5

woodland under forestry used

28 922

30.6

27 982

29.6

- 940

-3.3

shrub/grassland under forestry

0 767

0.8

1 518

1.6

751

97.9

unused woodland

3 783

4.0

3 468

3.7

- 315

-8.3

unused shrubland

4 710

5.0

4 423

4.7

- 287

-6.1

unused grassland

1 237

1.3

1 201

1.3

-36

-2.9

unused bare land

2 221

2.4

1 952

2.1

- 269

-12.1

wetland

3 277

3.5

3 668

3.9

391

11.9

3 607

3.8

3 622

3.8

15

0.4

arable land permanent crops pastures and meadows

water bodies

Table 5 -Evolution of specific land use/cover between 2001 and 2003

7.2.2.

Environmental parameters

A lot of ancillary information on environmental parameters were gathered on the ground like presence of erosion, irrigation, isolated trees, noise, natural hazards, transect and landscape photos. Experience from the surveys showed that gathering this kind of variables was extremely related to human perception and time of observation and therefore very difficult to compare between surveyors. The surveyed variables have therefore to be simplified in order to focus mainly on land use and land cover and the use of photos should be generalised for registering and checking the crop and its state of growth.

8.

CONCLUSIONS

The pilot surveys carried out in the Member States in the period 2001-2003 demonstrated the feasibility of the LUCAS project at a Community scale. During this first phase, it was shown that the project could:



provide a common methodology and classification for data collection and estimates of land use and land cover;



provide harmonised information for the entire territory of the European Union;



potentially yield early information on the areas under crops;



be used as a platform for dedicated surveys on soil erosion, the landscape, etc.



provide statistical data necessary for implementing some indicators measuring the integration of environmental concerns in the common agricultural policy.

For 2006, work has to be continued in the following areas:



Implementation of the 2006 survey in order to allow confirmation of the preliminary results emphasised in the Greek and in the 2005 pre-pilot surveys.



Calculation of the efficiency of the methodology with some preliminary work on the post-stratification of the Polish regions surveyed in 2005.



Coordination of the LUCAS 2006 survey with the 2006 Corine Updating process.



Further methodological analysis of 2001/2003 data and their integration with the 2006 dataset.

9.

REFERENCES

Bettio M., Delincé J, Bruyas P, Croi W,Eiden G: Area frame surveys: Aim, Principals and Operational Surveys Bruyas, P., Kayadjanian, M., Vidal, C. 2. Results of LUCAS survey 2001 on Land Use in Building Agri-Environmental Indicators. Bruyas P., Kayadjanian M., Vidal C.: The uses of European land and their spatial distribution BRUYAS P.: Land Use-Land Cover : LUCAS 20011 Primary Results Consorzio ITA: Report on ICARE – Integrated Crop Area Estimate Delincé, J.: Outline of project LUCAS Delincé J. : A European approach to area frame survey. Deumlich D. , ZALF: Analysis of LUCAS data on soil erosion KayadjanianM., Bettio B.: LUCAS - Monitoring EU territory using an area frame sampling approach Eiden G., Bettio M.– LUCAS – A EU-wide land use/land cover area frame statistical survey : Data Exploitation Towards Agro-environmental Indicators Eiden G., Jacques P., Theis R.: Linear landscape features in the European Union Developing indicators related to linear landscape features based on LUCAS transect data Eiden, G., Kayadjanian, M., Vidal, C. : Capturing landscape structures: tools. In: From land cover to landscape diversity in the European Union European Commission: The Lucas survey. European statisticians monitor territory. EUROSTAT (2000): Manual of Concepts on Land Cover and land Use Information Systems. Theme 5: Agriculture and Fisheries: Methods and Nomenclatures. Gallego, F.J.: Maximising distances between replicates in stratified systematic sampling Gallego, F.J. : Sampling Frames of Square Segments. GeoApikonisis Ltd: Methodological Analysis of the results of the LUCAS 2001 JRC ISPRA: Report on the Greek Agricultural Survey Optimisation KUL: Report on the Landscape Indicator Project Landsis g.e.i.e: Report on the Exploitation of data from the Community's LUCAS survey

Smile Life

When life gives you a hundred reasons to cry, show life that you have a thousand reasons to smile

Get in touch

© Copyright 2015 - 2024 PDFFOX.COM - All rights reserved.