Presentation is loading. Please wait.

Presentation is loading. Please wait.

LUCAS 2009/2012 results Optimising comparability and Quality Reports Item 4.1 Working Group for Land Cover and Land Use Statistics 2015.05.11-12 Bech.

Similar presentations


Presentation on theme: "LUCAS 2009/2012 results Optimising comparability and Quality Reports Item 4.1 Working Group for Land Cover and Land Use Statistics 2015.05.11-12 Bech."— Presentation transcript:

1 LUCAS 2009/2012 results Optimising comparability and Quality Reports Item 4.1
Working Group for Land Cover and Land Use Statistics Bech building, Quetelet (ESTAT.E4)

2 LUCAS sample: two phase design
Stable since 2006 Stratification Aiming at NUTS2 level estimates New regions added in each survey Fine-tuning in methodology (experience and feedbacks) Trade-off: improvements versus ensuring comparability Improve comparability of LUCAS results

3 LUCAS 2009 and 2012 results Optimising comparability 3 WPs
Improvement in the classification Area sampling applied with differences (thresholds and allocation) 3 WPs proposing proper imputation models for the areas not covered (e.g. using LUCAS 2009 observations for some areas not covered in LUCAS 2012). QRs New statistical tables

4 Classification change: Land Cover
Harmonised criteria of coverage (10%) More restrictive definition of Bare-land (from a coverage of 50% to 90%) Exclusion of mire and swamp forests from land cover peatbogs and the contextual assignment of points to woodland In comparison with 2009, this explains mostly the decrease of Bare-land, due to the more restrictive definition and the swap from Wetland to Wooded areas.

5 Classification change: Land Use
2 classes suppressed (added parameter status) Hunting Natural reserve the suppression caused a redistribution of the areas of the different uses and impacted the comparison with previous year.

6 LUCAS 2009 and 2012: Exclusion criteria from second phase sample

7 PROPOSED APPROACH Separate treatment of the different modules
Combining estimation with modelling 2009 only elevation 2012: elevation + accessibility < 1000m > 1500m

8

9 Weighting factor [direct and calibrated estimator]
Direct: inverse of selection probability Post – Calibration Elevation class: m m m >900 NUTS2¦¦strata NUTS1¦¦elevation NUTS0¦¦strata¦¦elevation 531

10 Main issues addressed

11 Results 8 (total area)+ 8 (area < =1500) results
For each year, Starting from different input data set: Base and "projected w/t-1, t observation" Direct and calibration estimator Same procedure excluding from points > 1500 [open issue: how to estimate area above 1500 m; is this kind of data vailable at national level] 8 (total area)+ 8 (area < =1500) results Ongoing assessment of best approach

12

13 Combining different options

14 LUCAS old results: comparing national sources - Estonia

15 LUCAS revised results: comparing national sources - Estonia

16 Land cover by country, % of total area
Land use and Land cover, 2012 Land Use in EU, % of Total area Land Cover in EU, % of Total area Land cover by country, % of total area

17 LUCAS Primary data: 2006,2009,2012 Country-specific files available at

18 LUCAS 2009 in-situ micro-data
Projected/imputed

19 LUCAS 2012 in-situ micro-data
Projected/imputed

20 Your feedback Best combination of weighting method Results: Two weeks
Critical Classes Geographical level Two weeks


Download ppt "LUCAS 2009/2012 results Optimising comparability and Quality Reports Item 4.1 Working Group for Land Cover and Land Use Statistics 2015.05.11-12 Bech."

Similar presentations


Ads by Google