1. Preliminary results and experiences of the ETS
in Sweden, Flanders, Denmark,
and The Netherlands
Follow Up from Copenhagen 2010
Joint Presentation
Flanders
Sweden
The Netherlands
Denmark
2. Goal of the presentation
Share experiences
Generate discussion
Illustrate issues/problems related to the LPIS
QAF that affect us all and should be addressed
Address specific issues related to the different
implementations of LPIS
The State of LPIS
Denmark: performed a total refresh in 2006/2007
After update implemented a continuous update cycle
Sweden: performed a total refresh in 2008
After update implemented a continuous update cycle
The Netherlands: performed a total refresh in 2009
After update implemented a continuous update cycle
Flanders has a continuous update strategy since
2004
3. 1. General Remarks LPIS QAF
2. General Experiences
3. Quality Elements 1, 2, 3, 4 and 5 (sample)
1. 3 Different LPIS implementations
2. Results (before and after changes Copenhagen)
3. General remarks on results
4. Specific issues related to the different implementations of LPIS
5. Impact latest changes LPIS QAF
4. Quality Elements 6 and 7 (IACS)
1. Results
2. General remarks on results
5. Conclusions/Suggestions
Structure/Set Up
General
The QAF for LPIS improves quality
awareness
The QAF, especially the ETS, is
elaborate and complex to perform
Introducing waivers for different
implementations of LPIS improves ‘fit
for purpose’ of the ETS
This presentation will demonstrate that
meeting the thresholds will be a
challenge… even after a complete
refresh
1. General remarks
TAKING INTO ACCOUNT
THE TYPE OF SYSTEM
AND WORKFLOW
4. Tools for performing the ETS
Different approaches
FL, DK, and SE used standard functionality and programs
NL created a specific tool (add on in ArcGIS)
Why create a customized tool?
Guide the workflow
Standardize the performance of operators
Facilitate the analysis
Incorperating QE’s into regular quality processes
1. General remarks
There’s a flyer available detailing the functionality of the ETS-Tool (Alterra)
2. General experiences
LPIS QAF isnLPIS QAF isn’’t an of thet an of the
shelf productshelf product
‘‘TranslationTranslation’’ of instructions toof instructions to
national situationnational situation
Performing the ATS and ETSPerforming the ATS and ETS
After Copenhagen some issues were addressedAfter Copenhagen some issues were addressed
but it became even more complexbut it became even more complex
Subject to misinterpretation (e.g. calculations)Subject to misinterpretation (e.g. calculations)
Purpose unclearPurpose unclear
LPIS QAF still evolving/changingLPIS QAF still evolving/changing
LPIS QAF
5. 2. General experiences
Overview time (in hours) needed for QAF
Flanders Netherlands Sweden Average
2010
Preparations 560 240 400 400
Software development 80 325 80 162
ATS 160 400 120 227
ETS 240 430 280 317
Analyses and reporting 80 60 160 100
Total 1.120 1.455 1.040 1.205
Total per parcel 1.40 1.16 0.83 1.10
2011 and further expected
ETS 240 430 280 317
Preperations, analyses etc. 80 100 240 140
Total 320 530 520 457
Total per parcel 0.40 0.42 0.42 0.42
Previous estimates from the JRC/EC were exceeded by 4 timesPrevious estimates from the JRC/EC were exceeded by 4 times
Demands substantial effort from MSDemands substantial effort from MS’’s both in time and in costs.s both in time and in costs.
2. General experiences
Overview costs for LPIS QAF
Flanders Netherlands Sweden Average
2010
Preparations € 28.000.00 € 21.720.00 € 20.000.00 € 23.240.00
Software development € 4.000.00 € 29.440.00 € 4.000.00 € 12.480.00
ATS € 8.000.00 € 36.200.00 € 6.000.00 € 16.733.33
ETS € 9.000.00 € 38.915.00 € 14.000.00 € 20.638.33
Analyses and reporting € 4.000.00 € 5.430.00 € 8.000.00 € 5.810.00
Total € 53.000.00 € 131.705.00 € 52.000.00 € 78.901.67
Cost per parcel € 66.25 € 105.36 € 41.60 € 71.73
2011 and further expected
ETS € 9.000.00 € 38.915.00 € 10.500.00 € 19.471.67
Preperations, analyses etc. € 4.000.00 € 9.050.00 € 12.000.00 € 8.350.00
Total € 13.000.00 € 47.965.00 € 22.500.00 € 27.821.67
Total per parcel € 16.25 € 38.37 € 18.00 € 25.29
Costs differ between member states but on a whole are
acceptable provided the LPIS QAF becomes really effective
6. From the percespective of
3 different LPIS implementations
Flanders : Agricultural Parcel (AP)
Sweden: Farmers Block (FP)
Denmark & The Netherlands: Physical Block (PB)
3. QA elements 1,2, 3, 4 and 5 (QC sample)
• reflects the real land-use as cultivated by farmers
• relatively high effort to maintain due to changes in land-use of farmers
• created and maintained based on farmers input and digital orthoimagery and other
available datasources
Creation of a Land Parcel Identification System (LPIS):
The agricultural parcel system
3.1 Three Different LPIS implementations
7. • relatively stable borders (streets, rivers, ...)
• only one farmer with n (farmers) parcels per reference parcel
• created and maintained based on farmers input and digital orthoimagery and other
available datasources
Creation of a Land Parcel Identification system (LPIS):
The farmers block system (Ilot)
3.1 Three Different LPIS implementations
• relatively stable borders (streets, rivers, ...)
• allows n farmers and m (farmers) parcels inside a reference parcel
• created and maintained based on farmers input and digital orthoimagery and other
available datasources
Creation of a Land Parcel Identification system (LPIS):
The physical block system
3.1 Three Different LPIS implementations
8. 3.2 Preliminary results
One Reference
parcel…
temporarily four
LCC-areas
What is expected of us in the ETS?
3.2 Preliminary results
Quality Element - 1
COMPARING THE SUM OF THE REFERENCE AREA
WITH THE SUM OF THE OBSERVED AREA
PATH
Reference = 2,11 ha
Observed (eligible) = 1,98 ha
6,2% difference
9. FL
(AP)
SE
(FB)
NL
(PB)
DK
(PB)
JRC/EC
RP sample sizeRP sample size 800800 12501250 12501250 800800 N/AN/A
CheckedChecked 276276 415415 287287 923 N/AN/A
SkippedSkipped 2424 88 1212 83 N/AN/A
Quality Element 1Quality Element 1 ––
ResultResult
100,3100,3 99,599,5 98,698,6 100,4 98%98% -- 102%102%
Maximum eligible area:
maximum eligible area is the recorded
quantity of land for which farmers can,
from a landcover perspective, apply for
aid under SAPS / SPS
Quality Element nr. 1
3.2 Preliminary results
BEFORE/AFTER COPENHAGEN
√ √ √ √
NO PROBLEM WITH QE-1
3.2 Preliminary results
EXAMPLE SKIPPED REFERENCE PARCEL
10. 3.2 Preliminary results
Quality Element – 2
Proportion of non conforming RP’s
Three groups related non-coforming:
• the reference parcel holds an unwaivered potential critical defect
• the difference between the eligible area observed and recorded
exceeds the threshold
• the LUI contains unwaivered contaminations of ineligible area
Quality Element nr. 2
Proportion of RPs with incorrect
recorded area
FLFL
(AP)(AP)
SESE
(FB)(FB)
NLNL
(PB)(PB)
DKDK
(PB)(PB)
JRC/ECJRC/EC
RP sample sizeRP sample size 800800 12501250 12501250 800800 N/AN/A
DigitizedDigitized 276276 312312 287287 800800 N/AN/A
Non ConformingNon Conforming 133 140 41 255 N/AN/A
Quality Element 2Quality Element 2 ––
ResultResult
48%48% 41,3%41,3% 15,1%15,1% 31,8%31,8% <= 5%<= 5%
BEFORE COPENHAGEN
3.2 Preliminary results
X X X X
NOT COMPLYING!
11. 3.2 Preliminary results
Quality Element nr. 2
RP – size taken into consideration
For Reference parcels greater than 5000m2
97 % and 103 %
For Reference parcels between (or equal to) 2000m2 and 5000 m2.
95 % and 105 %
For Reference parcels less than 2000 m2
93 % and 107 %
Modified after Copenhagen
Quality Element nr. 2
Proportion of RPs with incorrect
recorded area (taking into account the
RP size)
AFTER COPENHAGEN
3.2 Preliminary results
FL
(AP)
SE
(FB)
NL
(PB)
DK
(PB)
JRC/EC
RP sample sizeRP sample size 800800 12501250 12501250 800800 N/AN/A
CheckedChecked 210210 415415 287287 923923 N/AN/A
Non conformingNon conforming 29 123 40 255 N/AN/A
Quality Element 2Quality Element 2 ––
ResultResult
13,8%13,8% 39,4%39,4% 15,1%15,1% 31,8%31,8% <= 5%<= 5%
X X X X
STILL NOT COMPLYING!
12. Quality Element nr. 2
Distribution of RPs, according to
the correctness of the eligible
area recorded
FL
(AP)
SE
(FB)
NL
(PB)
DK
(PB)
% RPs% RPs % RPs% RPs % RPs% RPs % RPs% RPs
00‐‐2%2% 71,171,1 42,642,6 76,276,2 57,557,5
22‐‐4%4% 15,615,6 21,321,3 12,112,1 2121
44‐‐8%8% 5,25,2 20,020,0 8,38,3 12,112,1
88‐‐12%12% 2,82,8 5,65,6 1,91,9 4,44,4
1212‐‐20%20% 2,82,8 6,76,7 0,80,8 2,32,3
2020‐‐50%50% 1,41,4 3,63,6 00 1,81,8
>50%>50% 0,50,5 0,30,3 0,50,5 1,01,0
TotalTotal 100100 100100 100100 100100
BEFORE & AFTER COPENHAGEN
3.2 Preliminary results
Bulk of the deviations
between 0 – 8%
Purely informative!
Quality Element nr. 3
Categorization of the RP with
incorrect recorded area
BEFORE AND AFTER COPENHAGEN
Causes for non conformity FL (AP) SE (FB) NL (PB) DK (PB)
Changes in the land not
applied
30 82 8 40
Revisions of Regultion not
applied
0 8 0 1
Incomplete processsing 0 0 7 0
Erroneous processing 0 64 0 211
Incompatible LPIS design 0 0 12 0
Other 0 0 15 3
Acceptance number 14 15 13 40
3.2 Preliminary results
NOT COMPLYING
ON CERTAIN POINTS!
13. 3.2 Preliminary results
An example a of Critical Defect
Is not adhering to the LPIS implementation (in this case
the block paradigm) truly a (potential) critical defect??
Quality Element nr. 4
Occurrence of RP with potential
critical defects
BEFORE COPENHAGEN
FL SE NL DK
Number of (Potential) Critical
Defects
57 95 10 63
Acceptance level 3 3 3 10
3.2 Results
(before and after changes Copenhagen)
Hier aangeven of
voldaan wordt aan
norm of niet….
X X X X
NOT COMPLYING!
14. Quality Element nr. 4
Occurrence of RP with
unwaivered critical defects
AFTER COPENHAGEN
INTRODUCTION OF SO CALLED WAIVERS
3.2 Results
(before and after changes Copenhagen)
FL SE NL DK
Number of Critical Defects 1 25 10 63
Acceptance level 3 3 3 10
√ X X X
FLANDERS SEEMS TO
BENEFIT THE MOST
OTHERS STILL NOT COMPLYING!
No waivers for the
Physical Block System
3.2 Results
Example of a waiver
15. Quality Element nr. 5
BEFORE AND AFTER COPENHAGEN
Ratio of declared area in relation
to the maximum eligible area
inside the reference parcels
FLFL SESE NLNL DKDK
90 – 110% 142142 261261 NANA NANA
Outside 90 ‐ 110% 2424 5151 NANA NANA
Acceptance number 1212 1717 NANA NANA
3.2 Results
(before and after changes Copenhagen)
X X
Doesn’t take into account the underdeclaration
Farmers behaviour
Not quality LPIS
3.3 Problems encountered
Quantification on different scales
(VHR imagery vs data used to create LPIS)
Orthoimagery 2010
PATH
Orthophoto 2009
16. 3.3 Problems encountered
To be Eligible or not to be Eligible
Orthophoto 2009
Orthophoto 2008 Orthoimage 2010
6655-628-4546
• LPIS recorded area 1,27 ha (blue)
• ETS observed area 1,10 ha (red)
• 86,3% = non-conforming
3.3 Problems encountered
17. 6655-628-4546
• LPIS recorded area 1,27 ha (blue)
• ETS observed area 1,10 ha (red)
• 86,3% = non-conforming
3.3 Problems encountered
• Recorded area (blue line)
• Observed area (red line)
3.3 Problems encountered
18. Declared Area
NOT CLAIMED
doesn’t mean
NOT AGRICULTURE
3.4 General Remarks on Results
3.4 General Remarks on Results
2009 DECLARED
2010 NOT DECLARED
(leased by a ‘non farmer’)
NEGATIVE IMPACT ON QE5 EVENTHOUGH
IT’S STILL AGRICULTURAL LAND!
19. 3.4 General Remarks on Results
Declared Area = 1,80
Observed Area = 2,00
Percentage 90% declared
PARTIALLY NOT DECLARED
NO RISK FOR THE FUND!
OBSERVED
REFERENCE
Potential critical defect
Reference parcels
established by a field visit
should have a waiver
example:
3.4 General Remarks on Results
20. 6238-457-6885
• LPIS recorded area 1,73 ha (blue)
• ETS not digitized as boundary not
possible to draw
• Potential critical defect,
less than 100% eligible
• Reference parcel eligibility and
boundary established in field visit
• Is should NOT be a critical defect
3.4 General Remarks on Results
6238-457-6885
• LPIS recorded area 1,73 ha (blue)
• ETS not digitized as boundary not
possible to draw
• Potential critical defect,
less than 100% eligible
• Reference parcel eligibility and
boundary established in field visit
• Is should NOT be a critical defect
3.4 General Remarks on Results
21. 4.1 Quality Element 6 & 7
(Flanders)
Not relevant for a system based on Agricultural Parcels
Farmers are obliged to indicate changes on a yearly bases
Same applies if farmers are supplied with a recent orthophoto
at the beginning of the campaign
Area not found for random OTSC: 47,64 ha
0,61% of the declared area
OK! Conformance level 2 %
Quality Element nr. 6 LPIS cumulative land changes
Quality Element nr. 7
On the spot check (OTSC)
rate of irregularities
5. Suggestions/Conclusions
General Remarks:
ETS should (together with the member states) be evaluated
Are all criteria necessary/meaningful?
Some QE’s will by very hard to meet with the current instructions and
required conformance levels
Purpose of the ETS is unclear
are we measuring the quality of the data presented to the farmer
at the beginning of the campaign?
measuring the risk for the fund?
More information about the rational of the certain ETS-instructions will
have a positive impact on the performance of the ETS
Current ETS doesn’t take into account the results of cross checks and
farmers input
22. 5. Suggestions/Conclusions
Suggestion:
Earlier this year with formed a group consisting of 6 member states/regions
In order:
to help each other
to clarify issues
where possible act as one
encourage different ways of looking at the same problem
not trying to reinvent the wheel twice……
This initiative is supported by a Dutch Reserach Institute
(Alterra, Wageningen UR) complementing on a scientific basis
We feel that this initiative has been succesful up til now. Some examples:
Joint presentations
Developing concepts related to GeoCAP
Facilitating exchange of information (www.geocap.wur.nl)
From our point of view the current LPIS QAF
• is still a draft version.
• following a prototyping approach.
(‘With us MSs in the role as guinea pigs….’ )
To get out of this role we suggest in order also to help the
JRC/EC and more efficiently and more effectively establish
the QAF in the MS’s we consider
• An advisory board consisting of representatives from
MS (comparable to SDIC’s in INSPIRE) as a necessary
instrument
5. Suggestions/Conclusions
23. 5. Suggestions/Conclusions
• Quality Element 1
• Good indicator for the LPIS Quality
• Quality Element 4
• Waiver for RP’s that were checked on the spot
• Quality Element 5 only those parcels are relevant where:
• Observed Area < Delcared Area < Reference Area
• Quality Element 6 not relevant:
• for a system based on Agricultural Parcels
• if recent orthophoto’s are provided to the farmers (each year)
Questions/Discussion