the eagle inspire project land cover lena hallin-pihlatie / finnish environment institute christian...

Post on 18-Jan-2016

220 Views

Category:

Documents

1 Downloads

Preview:

Click to see full reader

TRANSCRIPT

The EAGLE INSPIRE project Land Cover

Lena Hallin-Pihlatie / Finnish Environment Institute

Christian Ansorge / EEA

Gebhard Banko / UBA

NRC EIS workshop, EEA

25-26 November 2015

2

● The EAGLE group

● Focus of the ”EAGLE INSPIRE project”

● Project set-up○ People○ Tools○ WPs○ Deliverables

● Recommendations and conclusions

Overview

3

● Eionet Action Group on Land Monitoring in Europe (EAGLE)

○ Representatives of National Reference Centres (NRC) for Land Cover in the EEA´s EIONET network(European Environmental Information and Observation Network)

○ Other Land Monitoring experts from mapping agencies, universities and the private sector

● Established in 2009 as a self-initiative without external funding

○ Open „membership“ based on own commitment

● FP7 Harmonised European Land Monitoring (2011-2013)

● EEA Call for Tender in 2013 -> EAGLE Consortium○ Several small projects from 10/2014-9/2015○ Contracts with EEA ○ Part of the Copernicus Land activities

What kind of EAGLE?

4

● Contract with EEA• Time frame: March 2015 - September 2015

● Task: To provide EEA with transformation rules and validated sample datasets for Corine Land Cover (CLC) and Urban Atlas (UA)○ According to the INSPIRE Annex II: land cover, vector data

model (LCV)• Not part of the contract: Transformation of the whole CLC / UA data sets

● People○ Lead: Austrian Environment Agency (UBA)○ Co-Lead: SYKE○ Partners: ALTERRA, IGN Spain, DGT Portugal○ Reviewers: Spain (Uni Malaga & Barcelona), ReSAC Bulgaria○ EEA: Christian Ansorge, Ana Sousa

• Montly TeleCons with EEA

The EAGLE INSPIRE project

5

Focus on EEA Land Cover products

First step to support the development of a process with which EEA can provide Corine Land Cover (CLC) and Urban Atlas (UA) datasets in an INSPIRE conformant way

Corine Land Cover

http://land.copernicus.eu/pan-european/corine-land-cover

Urban Atlas

http://land.copernicus.eu/local/urban-atlas

6

Project set-up: WPs, tools and processes

WP1 Mapping to INSPIRE• Mapping of

attributes from source data to target data structure

WP2 Transformation to GML 3.2.1• Setting up and

testing tranformation rules

• ETL process

WP3 Validation and testing• Setting up

validation tests• Validation of GML

to ensure provision of valid GMLs

7

WP 1 Mapping to INSPIRE

WP1 Mapping to INSPIRE• Mapping of

attributes from source data to target data structure

● Goal○ Human-readable transformation rules○ To map from Source data model Target data model○ EXCEL-file

● Templates○ JRC http://inspire.ec.europa.eu/data-model/approved/r4618/mapping/ ○ ESDIN www.esdin.eu (used in ELF)○ CDDA template

● Evaluation○ Decision to develop the Matching table template & methodology

further to better support• Mapping of attributes at a more detailed level • The communication/documentation between WP1, WP2 and WP3

8

Source Data Model: CLC

Code IDGeometry

Target Data model: Application schema

9„IR Obligation“ „Voluntary“

Vector Raster

Project

10

Official INSPIRE Mapping Tables

11

Download Empty Mapping Table

● Download

Download File as *.xml

12

LCV Mapping Table in Excel: .xml .xls

13

Complex Data Type Example: Extent

Complex data typescannot be mapped only by one row

14

● Based on JRC mapping table and on matching tables of the ESDIN, ELF and CDDA projects

● Feature types and data types included and complex data types opened in the same sheet○ LandCoverVector○ LandCoverNomenclature○ gmlBase, Feature and baseType2

• Exception: AddressRepresentation

● Extra columns for communication purposes

● Separate Readme, changeLog and explanatory ”code list” sheets

Our matching table approach

15

Matching table template CLC: main sheet

16

Simplification made

● Address○ It is mandatory to provide the address of the organisation, which is responsible for the

code list (ResponsibleParty>RelatedParty>Contact)○ Reference to the INSPIRE Annex I theme: address (ad)○ Added as xml snippet to the Matching table

17

Example of values filled in Mapping table

● Land Cover Dataset○ Including void reason values

○ INSPIRE Void Reason Values:

Final namespace structure has not been

decided upon

inspireId

18

● The URI to the code list containing the Land Cover classes has to be provided

Land Cover Nomenclature

● Code list for CLC: http://dd.eionet.europa.eu/vocabulary/landcover/clc

● Information about the organisation in charge has to be given through the RelatedParty attribute

19

● The Land Cover Class Value is provided as a persistent URL

● For example for CLC class 112:

http://dd.eionet.europa.eu/vocabulary/landcover/clc/112

LandCoverObservation

20

Example GML for LandCoverUnit

Code

ID

Geometry

21

● Goal○ Machine-readable transformation rules○ INSPIRE datasets should be delivered in GML

3.2.1

● Transformation tool: HALE or FME?o FME was chosen mainly due to preferences of

partners

● An Extract, Transform and Load (ETL) process was designed in FME Workbench: UA and CLC○ Sample datasets:

• CLC polygons from 3 countries (10 per CLC class)• UA data set from Paris

WP 2 Transformation to GML

WP2 Transformation to GML 3.2.1• Setting up and

testing tranformation rules

• ETL process

22

● LandCoverUnit are member elements inside of the LandCoverDataset

● The structure cannot be reproduced in FME

● Solution approved by EEA○ Slight modification of the

LandCoverVector.xsd○ The definition of the attribute

’member’ was altered

Association role ’Member’

23

● Step-wise manner○ 1. Validation against the

Target Schema (LandCoverVector.xsd)

○ 2. GML schematron validation

○ 3. Theme-specific schematron validation

● Cooperation with eEnvplus project

WP 3 Validation of GMLs

Validation methodology (Tracasa, 2014)

WP3 Validation and testing• Setting up

validation tests• Validation of GML

to ensure provision of valid GMLs

Conformance tests of the validation

24

Done

Out of project scope

25

Validation example

● Check of the content of the lcv:observationDate element

○ Wrong: „2013-12-01T:00:00:00+01:00“○ Right: „2013-12-01T00:00:00+01:00“

26

Encoding of Coordinate System

● Three possible ways to encode:○ EPSG:3035○ http://www.opengis.net/def/crs/EPSG/0/3035○ urn:ogc:def:crs:EPSG::3035

● Valid in GML 3.2.1• urn:ogc:def:crs:EPSG::3035

○ Example: <gml:Surface gml:id="EU-1804437-1" srsName="urn:ogc:def:crs:EPSG::3035">

27

Issues related to Land Cover Nomenclature

● Constraint: The embedded description or the external description shall be provided.

1) External Description attribute

- The name and the URI to an external document has to be provided

2) Embedded Description attribute- The classification described by land cover meta language LCML,

ISO 19144-2)

Embedded description There is not an official schema available

Type not specified (Type:any)

Not possible to validate

External description The „external description“ attribute was called „online

description“ in previous versions of the TG LC

For each LC class a code, a name, a definition and an RGB value has to be provided

Typo - correction

needed in TG LC and IR

28

● All three steps (WP1, WP2, WP3) are documented as part of a project report

● Matching tables○ Matching table template, UA Matching table and CLC

Matching table

● Separate FME Workbench for CLC and UA● Tutorial video to explain the structure of the FME Workbench

● Schematron file used for validation

Project deliverables

29

● Repository for○ Matching tables○ Schematron rules

● To fix typos and to discuss and resolve, for example○ LC related, for example:

• onlineDescription ->externalDescription • issue related to embeddedDescription• Associsation role ’Member’

○ General, for example:• encoding of voidable attributes of multiplicity 0..1 or 0...*• nilReason• Xsi:nil=’true’

● Report problems to the software companies○ FME issues ○ GIS clients should adapt to INSPIRE GML

● Share project results

Project recommendations

30

● The TG LC, the IR, the xml schemas contain typos and inconsistencies

● The software tools do not always make your life easier…

However,

● Exponential learning curve… gives hope

● Collaborative projects - > success

● There is a framework for addressing and resolving issues (MIG, Thematic Cluster platform)

Conclusions

top related