LPISQA Legacy/ETS/Definitions/Method

From Wikicap - European Commission

Consequences of sub-parcels, super-parcels and hybrid implementation

It can be demonstrated (see slide 24+ of this presentation) that sub-parcels and super-parcels (i.e. aggregated parcels) can easily deviate from the optimal representation of the land for the purpose of the CAP. However the optimal situation is not always feasible. Therefore it is worth checking whether a particular design doesn’t affect the LPIS QA results. More specifically

  • Sub-parcels may create smaller parcels, which may affect QE2 results
  • Super-parcels may make the declarations “fuzzy” may affect QE5 results
  • Super-parcels and hybrids (mismatching third party boundaries) cause potential critical defects and may affect QE3 results

NOTE: For implementing landscape features common to two neighbouring agricultural parcels it is worth considering “object referencing” cardinality (multiplicity), as:

  • a separate identification of the feature as a reference parcel may lead to declaration on the same land by two farmers (risk of a double declaration of that area).
  • Including the half of the feature to each of the bordering parcels will lead to invisible and thus arbitrary adjudication of the land.

Delineation and coding non-agriculture land cover in ETS inspection

Non-agriculture land cover has to be delineated only when it needs to be excluded from the inner area; i.e. when the patch of combined non-agriculture land exceeds 0.1 ha. This delineation represents "holes" in the polygon. Including smaller holes is allowed, but is not mandatory. The list and description of non-agriculture land cover classes relevant to the ETS are given in measure 10105 of Annex I where no LCCS codes are designated for these classes. On the other hand the abundance of non-agriculture land cover features should be reported according to the land cover type included in this list.

Why to delineate in ETS from scratch?

The purpose of the ETS is to verify whether the LIPS is fit for CAP purposes via a series of external and independent observations, NOT by "visual verifications". "Fit for purpose" is not equal to "meet the national LPIS specification". The mapping from scratch is an integral part of this independent external inspection and, furthermore, allows collecting more factual data that is probably stored in the LPIS. The possible "errors" induced by the poor quality and the orthoimage or incorrect interpretation are managed through the prior orthoimage checks and the “four-eye” control of the ETS inspection.

Is there a technical tolerance for the ETS inspection of the parcels?

There is NO technical tolerance applied during the ETS observations as there is no need to apply such technical tolerance, because:

  1. The process is not establishing whether it is measuring the same land object or not. It IS SUPPOSED to measure the same land object, albeit at a different time.
  2. The value for the major quantitative measure (“total eligible area”) is calculated from hundreds of parcels. Any unbiased measurement is expected to balance out the small differences in measurements inherent to any measurement methodology..
  3. The other quantitative measure that requires accurate area measurement (“categorization of reference parcels allowing payment undue on ineligible land or excluding agricultural land”) has a conformance–level (3%), well exceeding the CAPI measurement accuracies identified during the feasibility test.
  4. The acceptance numbers provided by ISO2859-2 imply some kind of practical tolerance. It allows many more individual parcels to fail the inspection than a simple proportion of the specified LQ would suggest.

The role of thresholds, waivers, and tolerances

The Commission services consider thresholds, waivers, and tolerances as methodological instruments that vindicate issues well before they enter the raw ETS-scoreboard. As these instruments prevent “reporting noise“, they are very useful. However they also produce too much “filtering”, which prevents the analysis of any true signal that comes with the raw observations.

How to inspect Landscape Features (LF)

LF can be taken as part of the eligible area in an agricultural parcel when is on the immediate border of the agricultural parcel.

LF can be modeled as land cover sub-parcels of a larger RP (e.g. a tree in middle of the crop). In LPIS QA LF that resides inside one and only one RP should be considered as full part of that RP. Therefore a single RP has to be presented for sampling, which corresponds to the merged LUI and MEA of the RP.

Any isolated LF, not bordering arable land, grassland or permanent crop, is ineligible and should be out of scope as their MEA should be zero.

Only a LF on the immediate border of two or more RP can be considered as a separate item for inspection (=LUI).


NOTE: the small LF, if considered as an ordinary RP, can dramatically increase the number of spatial parcels in the lot and become a significant proportion of items in the LPIS QA sample, which may distort representativeness.

Do we need to map the land cover when the LUI boundary was derived from the RP perimeter?

Yes and follow these steps:

  1. using the reference parcel perimeter as a substitute of the LUI external boundary has to be considered as a single mapping step, conditioned by the fact that no ineliglibe land cover feature is present inside the 5 meter buffer around the unidentifiable RP perimeter.
  2. after this step, the further detailing of agricultural land cover classes and exclusion of non-agricultural elements should be continued as if a normal "measurable" parcel was involved.

For detailed instructions, please look at this article.

Why to distinguish between over- and underestimation of the reference area?

Even though only the former implies a financial risk, the Commission services are also interested in the fitness of the LPIS to give the farmers a proper support for the declaration process and to give an indication of potential risk of the particular system. As the differences are reported in gross terms appropriate conclusions can be drawn.

Establishing area values for eligible landscape features represented in measure 10104_2 (QE 2)

The ETS inspection is based on inventory of the land cover through mapping, i.e.as seen on the orthoimage coverage. Any mapping is an abstraction of real world phenomena. Depending on the concrete users’ need the abstraction of the same real world phenomenon may result in different geometries such as point, line, or polygon. Most frequently land cover features are abstracted as polygons. However for rows of trees or single trees a line or point representation might be more appropriate. But this does not create any obstacle in determining the area of such LF.

The LFs subject to retention are defined in the national GAEC rules, together with their size, shape, width and texture. Therefore, the area to be allocated to landscape features represented as lines or points can be calculated “on-the-fly”, using the regulatory width or the radius attributed to the landscape features. Point 5.1.2 of ETS Annex III (see Downloads) gives some recommendations regarding the geometry types that can be used for the representation of the landscape features, as listed in Annex III of Council Regulation 2009R73.